#agentforce — Public Fediverse posts

6 Key Takeaways from TDX26

The landscape of software development is evolving at breakneck speed. The TDX 2026 main keynote made one thing clear: we have officially entered the agentic revolution. Time after time, we think we found the keys to the big gates of the agentic era, soon to find out we need to pivot yet once again.

Celebrating the 10th anniversary of the event and the broader Trailblazer movement, Salesforce leadership shared a bold new vision for how businesses build applications and operate their workflows. 

The Rise of the Agentic Enterprise with TDX26

Probabilistic AI workflows are fundamentally changing the era of deterministic software, where rigid inputs always yield the exact same outputs. Salesforce’s vision for the future is the Agentic Enterprise, a model where human workers and AI agents collaborate dynamically across every workflow to skyrocket productivity. Real-world impact is already here: Telecom giant Lumen is saving over 300 hours a week with Agentforce, ADP is streamlining HR services, and PenFed Credit Union is successfully handling 75% of its routine IT tasks autonomously.

A standout success story from the keynote was Engine, an agentic-first B2B travel platform. Using Agentforce Builder and Trailhead resources, they built an AI virtual assistant named Eva in just 12 days. I find this timeline hard to believe. Today, Eva manages 50% of their customer support queries. It instantly recalls booking information and frees up human representatives to focus on more complex, high-value problem-solving.

Headless 360: But What Does This Mean?

At the core of this transformation is Salesforce Headless 360, a highly composable system that utilizes Model Context Protocol (MCP) endpoints to function like “Legos” for enterprise developers. Instead of operating in disconnected silos, these systems empower agents to actively reach into diverse data environments and communicate openly. For developers looking to monetize their custom AI innovations, Salesforce also introduced the AgentExchange, supported by a brand-new $50 million investment fund to help Trailblazers build, share, and sell specialized agent skills.

Next-Generation Developer Tools

To empower modern builders, Salesforce showcased an impressive array of groundbreaking tools:

• Agentforce Vibes 2.0: Developers can now seamlessly use the Anthropic and OpenAI SDKs directly within their environments to write, test, and evaluate code agentically. This is the second generation Agentforce Vibes which was originally announced at Dreamforce 2025. Developers are free to choose the AI LLM they prefer when developing on this platform. In a full circle moment, CLI seems to have earned a seat on the center stage. 
• Agent Script: This single, versionable flat file ensures AI agents adhere to strict organizational guardrails. This adds a necessary layer of determinism and trust to how complex tasks are routed. I personally thought Salesforce announced this a while ago.
• Agentforce Experience Layer: In a massive leap for UI development, builders can now define a rich interface component just once and seamlessly deploy it across the web, mobile, Slack, Microsoft Teams, and even ChatGPT. This innovation natively pushes the brand experience directly into the user’s natural flow of work. This is an important announcement and signals a big shift in Salesforce development strategy.

6 Key Takeaways from TDX26

The topics highlighted in this announcement paint a clear picture of where Salesforce’s vision is heading. I also observed the absence of some products and platforms during the announcement.

Here are my takes on the implications of this vision, if it can be successfully delivered:

1. UI: User interface on Salesforce is deprioritized. In the following years, we should expect the tools and products that customize the Salesforce UI to gradually lose their importance. Will we spend hours after hours to refine dynamic page layouts? Probably not.
2. Channels: The new focus is to meet the users where they already are. If they are on Slack, developers will find a way to visualize the data on Slack. If they are on Teams, the users will see the data there.
3. Integrations: Every tool either becomes an MCP (or API) provider or client or both. AI will help developers connect these tools.
4. Salesforce Code/Low-Code: I counted the times Apex or Flow were mentioned in the keynote. Each appeared only once or twice. Instead, the main focus was enabling development using AI and mainstream coding platforms (React got its moment on the center stage). Please read my additional comments in the conclusion section.
5. Data: Data 360 remains in the center stage both as a tool and functionality. Salesforce positions its value as the provider and processor of meaningful data.
6. Hybrid Agents: Salesforce still highlights Agent Script as a deterministic way of processing transactions pointing to challenges in Agent accuracy. Salesforce recommends hybrid approaches between deterministic and probabilistic agents.

Prefer to learn by watching? Here is the exact same information in this post, but in video form. ⬇️

Additional Features and Functionalities

Here is a list of the remaining new features and functionalities announced at the TDX 2026 keynote:

• Agent Exchange & $50M Fund: A new marketplace ecosystem allowing developers to build, share, and monetize specialized agent skills and sub-agents, supported by a newly announced $50 million investment fund.
• Agentforce Labs & “Getting Started”: A new environment for testing cutting-edge AI innovations before they are generally available. The “Getting Started” feature lets developers provision an org, set up Agent Script, and deploy an agent in seconds using a single terminal command.
• Agent Builder: A low-code canvas view designed for business users to build and collaborate on agents using natural language. The latest evolution of the tool that has been around for years.
• Testing Center & A/B Testing API: A dedicated testing experience with custom scores, evaluators, and automated smoke testing to simulate real user scenarios and ensure agents follow guardrails. Salesforce also announced a new A/B Testing API to help continuously improve agent performance.
• Observability Tools (SCDM API & Slackbot Telemetry): A new OpenTelemetry (OTEL) compatible SCDM API provides deep visibility into agents in production. Slackbot natively integrates telemetry. This allows admins to paste a session trace ID and instantly analyze and improve agent interactions within their flow of work.
• Multi-Agent Orchestration (Agent Fabric): Now generally available, this capability uses Mulesoft Fabric (or Agent Fabric) to connect multiple agents within an org, enabling them to collaborate and pass off tasks -such as a service agent flagging an opportunity for a sales agent.
• Agentforce Voice, Mobile SDK, and Enhanced Chat v2: Expanding how users interact with agents, Salesforce launched Agentforce Voice (now actively powering their 1-800-NO-SOFTWARE help line), Enhanced Chat v2 for rich web interfaces, and the Agentforce Mobile SDK for embedding native agent UIs into mobile apps.

Please also note that, with this event, Agent topics have been renamed to subagents.

Conclusion

Ultimately, TDX 2026 highlighted a big shift in Salesforce development strategy. It was one of the boldest events I have seen in terms of setting a new direction for Salesforce’s development vision.

Whether Salesforce can execute effectively against this vision remains to be seen. Regulated industries impose real limits on where Agentforce can be deployed. The shift from user-based to MCP/API consumption-based licensing adds another layer of complexity. The skills gap in the ecosystem remains a challenge as well.

However, if Salesforce is successful, this shift would mean a very different way of working for most of us.

As I mentioned above, Flow and Apex were mentioned only once or twice in the keynote. When the Apex topic came up, the presenter said developers won’t be writing Apex by hand. The Engine demo included Agentforce Vibes writing a React UI. This new rhetoric challenges the importance of Salesforce proprietary code/low-code development platforms. In the next five to ten years, we should see folks questioning the future of Omnistudio, which is essentially a LWC editor, screen flows and even Apex.

It is important to note that we still need experienced and skilled professionals who can test and improve the outputs of these AI-powered development tools. I highly recommend that developers adopt AI-assisted development, and admins adopt higher development skills mastering flows and code to prepare for this future.

Explore related content:

No Browser or UI Required: Salesforce Headless 360 Explained

Top 7 Key Takeaways from Salesforce Dreamforce 2025

Our Must-See TDX 2026 Session Guide

Transform Customer Engagement with Agentforce Services

Agentforce services help businesses deliver smarter, faster, and more personalized customer support using AI-powered automation. Improve agent productivity, streamline workflows, and enhance customer experiences with intelligent service solutions designed for modern businesses.

Get in touch: https://achieva.ai/agentforce/

#Agentforce, #AI #Customer #Service, #Automation, #CRM, #Customer #Support, #Digital #Transformation

🚀 Smarsh just rolled out an AI front‑door built for regulated firms, letting 59% of users go self‑service. Powered by generative AI and Agentforce, it plugs into Salesforce to boost compliance and customer service without heavy IT overhead. Curious how AI agents can streamline regulated industries? Read on! #AIAgents #Agentforce #Compliance #SelfService

🔗 https://aidailypost.com/news/smarsh-launches-ai-front-door-regulated-firms-hits-59-selfservice

Salesforce Agentforce Development — Transforming Field Service & Sales Productivity 🚀

Unlock the full potential of your CRM with Salesforce Agentforce Development — designed to streamline field service operations and supercharge sales performance.

Explore how innovation meets productivity — and take your field service and sales teams to the next level.

https://achieva.ai/agentforce/

#Salesforce #Agentforce #FieldService #SalesProductivity #DigitalTransformation

https://www.fogolf.com/1163663/liv-golf-debuts-ai-fan-caddie-second-screen-companion/ LIV Golf debuts AI ‘Fan Caddie’ second-screen companion #Agentforce #AIAgents(AgenticAI) #ArtificialIntelligence(ai) #ConversationalAI #CustomerData #DataPlatform #DataVisualisation #DigitalTransformation #DirectToConsumer(D2C) #ECommerce(eCommerce) #FanExperience #GenerativeAI(GenAI) #Golf #GolfNews #Hospitality #Personalisation #PredictiveAnalytics #RealTimeData #Retail #salesforce #Semiconductors #Slack #SportsTechnology(SportsTech) #ticketing

https://www.fogolf.com/1163663/liv-golf-debuts-ai-fan-caddie-second-screen-companion/ LIV Golf debuts AI ‘Fan Caddie’ second-screen companion #Agentforce #AIAgents(AgenticAI) #ArtificialIntelligence(ai) #ConversationalAI #CustomerData #DataPlatform #DataVisualisation #DigitalTransformation #DirectToConsumer(D2C) #ECommerce(eCommerce) #FanExperience #GenerativeAI(GenAI) #Golf #GolfNews #Hospitality #Personalisation #PredictiveAnalytics #RealTimeData #Retail #salesforce #Semiconductors #Slack #SportsTechnology(SportsTech) #ticketing

https://www.fogolf.com/1163663/liv-golf-debuts-ai-fan-caddie-second-screen-companion/ LIV Golf debuts AI ‘Fan Caddie’ second-screen companion #Agentforce #AIAgents(AgenticAI) #ArtificialIntelligence(ai) #ConversationalAI #CustomerData #DataPlatform #DataVisualisation #DigitalTransformation #DirectToConsumer(D2C) #ECommerce(eCommerce) #FanExperience #GenerativeAI(GenAI) #Golf #GolfNews #Hospitality #Personalisation #PredictiveAnalytics #RealTimeData #Retail #salesforce #Semiconductors #Slack #SportsTechnology(SportsTech) #ticketing

https://www.fogolf.com/1163663/liv-golf-debuts-ai-fan-caddie-second-screen-companion/ LIV Golf debuts AI ‘Fan Caddie’ second-screen companion #Agentforce #AIAgents(AgenticAI) #ArtificialIntelligence(ai) #ConversationalAI #CustomerData #DataPlatform #DataVisualisation #DigitalTransformation #DirectToConsumer(D2C) #ECommerce(eCommerce) #FanExperience #GenerativeAI(GenAI) #Golf #GolfNews #Hospitality #Personalisation #PredictiveAnalytics #RealTimeData #Retail #salesforce #Semiconductors #Slack #SportsTechnology(SportsTech) #ticketing

Salesforce pulls back from LLMs, pivots Agentforce to deterministic automation

https://completeaitraining.com/news/salesforce-pulls-back-from-llms-pivots-agentforce-to/

#HackerNews #Salesforce #LLMs #Agentforce #Automation #TechNews

Rebranding Salesforce to Agentforce is perfect if your core demographic is enterprise executives who think 'AI' means their email sidebar got smarter. #Agentforce #AIFlavored 🫠

Salesforce shares dropped 30% amid AI concerns, but Q3 revenue rose 8.6% to $10.26B, and full-year guidance was raised to $41.45–41.55B. Its Agentforce AI platform and acquisitions support long-term growth potential.

#Salesforce #AI #Agentforce #Informatica #SaaS #TechStocks #CloudComputing #RevenueGrowth #MarcBenioff #SoftwareIndustry #TECHi

Read Full Article Here :- https://www.techi.com/salesforce-stock-hits-lowest-valuation-despite/

Salesforce’s Agentforce is cutting support load – it deflects 46% of advertiser cases by using AI‑driven real‑time observability and session tracing. John Thompson says the platform gives teams instant insight, freeing engineers for higher‑value work. Curious how this could reshape your support ops? Read the full story. #Salesforce #Agentforce #AIObservability #AdvertiserSupport

🔗 https://aidailypost.com/news/salesforce-agentforce-deflects-46-advertiser-support-cases

Salesforce’s Agentforce is cutting support load – it deflects 46% of advertiser cases by using AI‑driven real‑time observability and session tracing. John Thompson says the platform gives teams instant insight, freeing engineers for higher‑value work. Curious how this could reshape your support ops? Read the full story. #Salesforce #Agentforce #AIObservability #AdvertiserSupport

🔗 https://aidailypost.com/news/salesforce-agentforce-deflects-46-advertiser-support-cases

AI會終結電話客服中心嗎？ChatGPT總結為「人機協作」，但人類的看法不那麼樂觀

BBC News 中文 2025-11-05 07:55:00 CST

Authors, lawyer up.

https://www.salesforceben.com/salesforce-faces-lawsuit-from-authors-over-ai-training-data/?utm_source=hs_email&utm_medium=email&utm_content=120329875

#Salesforce #AgentForce

Top 7 Key Takeaways from Salesforce Dreamforce 2025

Salesforce Break reviewed the press releases and sessions coming out of Salesforce Dreamforce 2025, and prepared the ket takeaways in this post, so you don’t have to go thorough all the materials.

The biggest announcements for Salesforce at Dreamforce 2025 were centered on advancing the company’s foundational “Agentic Enterprise” vision through enhanced control, deeper context integration, and widespread collaboration tools.

The announced functionalities were more evolutionary than revolutionary.

Here is the list.

Top 7 Key Dreamforce Takeaways

1. Agentforce 360 Platform, New Agentforce Builder, and Agent Script

The cornerstone announcement was the launch of Agentforce 360, the latest version of the comprehensive platform designed to unify AI, trust, and data capabilities across all Salesforce products. Salesforce has completely reimagined the entire Customer 360 platform as Agentforce 360, ensuring that every app is now “agentic”. This platform emphasizes providing users with more control than ever over their AI systems. To make development accessible to a wider audience, including line-of-business leaders, IT teams, service, and sales teams, a brand new AgentForce Builder was introduced, featuring a radically simplified, clean, and beautiful interface built from the ground up.

This capability is powered by Agent Script, a new scripting language that exposes the reasoning engine and allows users to define deterministic chaining actions and conditional logic. Agent Script unlocks patterns needed for mission-critical use cases, blending fluid agentic reasoning with the certainty of rules-based control in a unified instruction set to ensure agents are predictable and stay “on track,” preventing costly unpredictability. Agent Script can be built at the topic level where previous LLM-based non-deterministic functionality produced unpredictable results.

In addition, Salesforce announced Slack as its future conversational interface. Several sessions demonstrated deeper integrations in action. Another major change of course was the ability to use external LLMs for the Atlas Reasoning Engine. I believe this demonstrates that Salesforce is positioning Agentforce more as an orchestrator and collaborator of agents and AI capabilities rather than competing to become the agent for the enterprise.

2. Agentforce Voice

Agentforce Voice extends the power of the Agentforce platform by allowing agents to talk, bringing AI capabilities directly to contact centers and 800 numbers. Businesses can now configure the voice, tone, and personality of the AI right inside the AgentForce Builder. The goal is to deliver a unified customer experience across all channels, providing a highly human-like and interruptible conversational flow. A critical feature of Agentforce Voice is ensuring a seamless transition when the AI needs to transfer a customer to a human agent; the human representative automatically receives the full transcript and context of the AI conversation, allowing them to pick up the experience precisely where the AI left off. This functionality is available GA October ’25.

3. Intelligent Context Processing (Data 360)

Intelligent Context Processing tackles one of the greatest challenges for AI agents: understanding and utilizing vast amounts of complex, unstructured data. Agents often struggle with content in rich formats, pictures, tables, and existing workflows, the accumulated wisdom of the company. These new tools, built into Data 360, interpret and index this data by analyzing and parsing complex content (such as product manuals containing charts and images). This allows agents to pull in the exact, correct context required to deliver accurate and rich responses at the precise moment it is needed.

Furthermore, Data 360 enhances governance across both structured and unstructured data. Using natural language, administrators can create policies, such as masking internal FedEx employee contact details within agent responses, ensuring the information provided is not only accurate but also appropriate for the customer. It is not clear to us whether this is solely a rename of the product called Data Cloud. It seems that way.

4. Agentforce Vibes

Salesforce launched Agentforce Vibes as a new product that lets trailblazers quickly and easily build apps, dashboards, automations, and flows. Users achieve this via vie coding, which involves providing a simple, natural language description of what they want the platform to build. The core innovation of Agentforce Vibes is its deep contextual understanding; it speaks “the language of the business,” including the organization’s data, relationships, customers, products, and security permissions. This contextual intelligence allows Agentforce Vibes to rapidly translate a descriptive idea into deployable, production-grade Salesforce metadata (such as a screen flow). This drastically reduces development time, saving what could amount to dozens of manual clicks inside a traditional flow builder. This effectively elevates the capabilities of every developer. Interesting tidbits: Developers can develop using the coding language of their choice, and there is a local LWC preview function that will be launched soon.

5. Slackbot

Salesforce unveiled Slackbot as a new personalized, AI-powered companion that boosts productivity directly within Slack. It will launch for General Availability (GA) in January and draws on each user’s unique organizational context, including conversations, files, workflows, and apps. The tool moves users beyond asking simple questions toward achieving complex, tangible outcomes. For example, a user can ask Slackbot to handle a multi-step process with one command. It can review deal status, find compliance files, and draft a customer email in the user’s tone. Slackbot can also create a prep document and calendar invite for key stakeholders automatically. Slackbot will be the home of AI capabilities within Slack, even for customers who don’t use Salesforce.

6. Support for Third-Party Agents in Slack (AgentExchange)

Salesforce affirmed its vision of Slack becoming the “home for all AI” by announcing support for all third-party AI agents, such as ChatGPT, Claude and Gemini. This transformation positions Slack as an agentic operating system where external agents can exist as collaborative “teammates” alongside human employees. To ensure these external agents can perform sophisticated reasoning, they are grounded in the company’s real-time knowledge and context via a real-time search API and an MCP server. This initiative allows Salesforce agents to work in conjunction with agents from other platforms. This coupled with the AI-assisted enterprise search capabilities of Slack empowers Slack admins and users to be more productive.

7. Agentforce Observability

Agentforce Observability was introduced to help monitor and scale digital work in the new agentic enterprise era. It serves as one control center for managers to monitor and improve agent team performance. The tool gives leaders visibility into KPIs like escalation and deflection rates using Tableau Next analytics.

Most importantly, it features Agent Insights, which acts as a performance review by scoring every single agent session. This scoring helps managers find and analyze poor-performing conversations to uncover root causes like process issues. It enables tuning of agent prompts and behaviors for consistent results. This management layer is essential since prompts and loops alone aren’t enough.

This was a major pain point with the clients. I am happy Salesforce is addressing it with this new functionality that will be available for most clients.

Conclusion

I personally found the announcements more evolutionary than revolutionary. It was not a strong Dreamforce in terms of new functionalities covered.

Adoption challenges and cleanup are still needed to make current products appealing. These announcements mark real progress for Salesforce.

[youtube https://www.youtube.com/watch?v=_S18LAXcBYQ?version=3&rel=1&showsearch=0&showinfo=1&iv_load_policy=1&fs=1&hl=en-US&autohide=2&wmode=transparent&w=640&h=360]

Explore related content:

Salesforce Ushers in the Age of the Agentic Enterprise at Dreamforce 2025

Dreamforce 2025: Standout Sessions Streaming on Salesforce+

Salesforce Winter ’26 Release: Comprehensive Overview of New Flow Features

Slack transforms into an agentic OS as ChatGPT joins the workspace

https://web.brid.gy/r/https://nerds.xyz/2025/10/slack-agentic-os-chatgpt/

Salesforce is pushing further into the AI space with Agentforce 360, a platform for building and deploying AI agents. The enterprise AI competition is fierce!

How do you see AI agents impacting customer service and operations?
#AI #SoftwareDevelopment #EnterpriseTech #Agentforce #Salesforce
https://techcrunch.com/2025/10/13/salesforce-announces-agentforce-360-as-enterprise-ai-competition-heats-up/

Salesforce announces Agentforce 360 as enterprise AI competition heats up

https://web.brid.gy/r/https://techcrunch.com/2025/10/13/salesforce-announces-agentforce-360-as-enterprise-ai-competition-heats-up/

Salesforce Enters ‘Vibe Coding’ Fray With Agentforce Vibes, an AI Agent for Enterprises

#AI #Salesforce #AICoding #VibeCoding #Agentforce #EnterpriseAI #DevTools

https://winbuzzer.com/2025/10/02/salesforce-enters-vibe-coding-fray-with-agentforce-vibes-an-ai-agent-for-enterprises-xcxwbn

Critical Salesforce Vulnerability ‘ForcedLeak’ Exposes AI Agent Risks in AgentForce https://thecyberexpress.com/forcedleak-agentforce-vulnerability/ #TheCyberExpressNews #Vulnerabilities #TheCyberExpress #FirewallDaily #AgentForce #ForcedLeak #Salesforce #CyberNews #CRMdata

ForcedLeak Flaw in Salesforce Agentforce AI Agent Exposed CRM Data https://hackread.com/forcedleak-salesforce-agentforce-ai-agent-crm-data/ #Cybersecurity #Vulnerability #AgentForce #ForcedLeak #Salesforce #Security #AIAgents #CRM #AI

Critical #ForcedLeak flaw in Salesforce’s Agentforce AI agent exposed CRM data to remote attackers.

Read: https://hackread.com/forcedleak-salesforce-agentforce-ai-agent-crm-data/

#CyberSecurity #Salesforce #Agentforce #Vulnerability #AI

🚀 Empower Your Team with Salesforce Agentforce

Boost productivity and collaboration with AI-powered automation. Streamline workflows, enhance customer engagement, and unlock smarter decision-making. Transform the way your team works today with Salesforce Agentforce.

Get in touch: https://www.damcogroup.com/salesforce/agentforce

#Salesforce #Agentforce #AI #Automation #FutureOfWork #Productivity

@Salesforce just released #Agentforce 3. What does this mean for #enterprises and the #EnSW market?

Is it ready for the enterprise or are additional iterations required?

Many questions.

Some answers.

https://zurl.co/huzLr

Salesforce Agentforce 3 brings visibility to AI agents https://www.artificialintelligence-news.com/news/salesforce-agentforce-3-brings-visibility-ai-agents/ #salesforce #agentforce #enterprise #agents #tech #ai #news #technology

AgentForce: The AI That Doesn’t Just Answer, It Acts - Sujith Abraham from SalesForce https://youtube.com/shorts/RAh0iKCD7jQ

Spotify: https://open.spotify.com/episode/37RG8Kr4XeQO4kKgYT2ZQm?si=wLcs6GDqS_ytqQ_uHulMzw
Instagram: https://www.instagram.com/reel/DIXk29RIiBW/?utm_source=ig_web_copy_link&igsh=MzRlODBiNWFlZA==
TikTok: https://www.tiktok.com/@analyseasia/video/7490145496170892552?is_from_webapp=1&sender_device=pc&web_id=7198703856800581122
Main Site: https://www.analyse.asia/how-agentforce-is-transforming-businesses-in-asean-with-sujith-abraham/
Full YouTube Video: https://youtu.be/6GgBo7CUshw?ref=analyse.asia
Newsletter Signup: https://www.analyse.asia/#/portal/signup
LinkedIn Page: https://www.linkedin.com/company/analyse-asia/

#salesforce #agentforce #generativeai #agenticai #aiagents

Why Salesforce? Here’s the Real Answer - Sujith Abraham from Salesforce explains https://youtube.com/shorts/ejypnXlbFII

Instagram: https://www.instagram.com/reel/DISUafGxXiB/?utm_source=ig_web_copy_link&igsh=MzRlODBiNWFlZA==
TikTok: https://www.tiktok.com/@analyseasia/video/7490144683662904594?is_from_webapp=1&sender_device=pc&web_id=7198703856800581122
Spotify: https://open.spotify.com/episode/37RG8Kr4XeQO4kKgYT2ZQm?si=wLcs6GDqS_ytqQ_uHulMzw
Main Site: https://www.analyse.asia/how-agentforce-is-transforming-businesses-in-asean-with-sujith-abraham/
Full YouTube Video: https://youtu.be/6GgBo7CUshw?ref=analyse.asia
Newsletter Signup: https://www.analyse.asia/#/portal/signup
LinkedIn Page: https://www.linkedin.com/company/analyse-asia/

#salesforce #agentforce #generativeai #agenticai #aiagents

What Every CEO Needs to Know About #GenerativeAI Speed - Sujith Abraham from Salesforce https://youtube.com/shorts/C34dlRpZClI

Spotify: https://open.spotify.com/episode/37RG8Kr4XeQO4kKgYT2ZQm?si=wLcs6GDqS_ytqQ_uHulMzw
Instagram: https://www.instagram.com/reel/DIPvr90B19A/?utm_source=ig_web_copy_link&igsh=MzRlODBiNWFlZA==
TikTok: https://www.tiktok.com/@analyseasia/video/7490143867363216658?is_from_webapp=1&sender_device=pc&web_id=7198703856800581122
Main Site: https://www.analyse.asia/how-agentforce-is-transforming-businesses-in-asean-with-sujith-abraham/
Full YouTube Video: https://youtu.be/6GgBo7CUshw?ref=analyse.asia
Newsletter Signup: https://www.analyse.asia/#/portal/signup
LinkedIn Page: https://www.linkedin.com/company/analyse-asia/

#salesforce #agentforce

This Is What Seperates #AgentForce Apart from Every Other AI Platform by Sujith Abraham from #Salesforce https://youtube.com/shorts/qOVKG2pj6MY

Spotify: https://open.spotify.com/episode/37RG8Kr4XeQO4kKgYT2ZQm?si=wLcs6GDqS_ytqQ_uHulMzw
Instagram: https://www.instagram.com/reel/DINK3ddBv3-/?utm_source=ig_web_copy_link&igsh=MzRlODBiNWFlZA==
TikTok: https://www.tiktok.com/@analyseasia/video/7490142799342390546?is_from_webapp=1&sender_device=pc&web_id=7198703856800581122
Main Site: https://www.analyse.asia/how-agentforce-is-transforming-businesses-in-asean-with-sujith-abraham/
Full YouTube Video: https://youtu.be/6GgBo7CUshw?ref=analyse.asia
Newsletter Signup: https://www.analyse.asia/#/portal/signup
LinkedIn Page: https://www.linkedin.com/company/analyse-asia/

#generativeai #aiagent

Salesforce Launches Agentforce for HR Service to Make Employee Support Easy for Everyone

#Salesforce #Agentforce #AI AIagents #HRTech #EmployeeExperience #HRAutomation #FutureOfWork #DigitalLabor #ServiceCloud #EnterpriseAI

https://winbuzzer.com/2025/05/06/salesforce-launches-agentforce-for-hr-service-to-make-employee-support-easy-for-everyone-xcxwbn/

Next Tuesday, I will be speaking at the London Salesforce Marketer Group.

We will be talking about #Agentforce agents and how #AI can help improve the life of marketers.

Be prepared to get hands-on. We will work together to generate ideas for possible agents that you can use in your business.

Grab your spot using the link below before it is too late!

https://trailblazercommunitygroups.com/events/details/salesforce-salesforce-marketer-group-london-united-kingdom-presents-spring-into-agentforce-for-marketing/

How To Build Flex and Field Generation Prompt Templates in the Prompt Builder

Prompt engineering is one of the most important skills of our era. With the speedy adoption of AI Large Language Models, the use of language became a very important skill again. One could argue that effective natural language writing will potentially be more important than coding in the future. Prompt writing is also critical for Salesforce Agentforce.

While writing in English – or any other language that the LLM understands – is important, we need to keep in mind how LLMs work so that we can design well-written instructions, establish guardrails, and ground our prompts to produce accurate results.

Writing prompts is not necessarily a Salesforce skill. This is now a universally sought-after skill that everyone needs to master.

Introduction to Prompt Builder

Prompt Builder is a tool to facilitate the creation of AI prompts within Salesforce, enabling users to automate responses based on specific data inputs. This tool is essential for reducing inefficiency when delivering results, thereby increasing operational efficiency across various business functions. 

Integration with Salesforce Flow and Apex

Prompt Builder is very well integrated with Salesforce Flows and Apex. This integration allows for excellent customization and functionality expansion, enabling users to execute sophisticated prompts through elaborate workflows and automate intricate tasks. Additionally, it facilitates simple data retrieval from multiple sources, including Salesforce’s Data Cloud and external APIs.

Salesforce Prompt Template Types

Here are all Salesforce Prompt Template Types:

1. Einstein AI-Generated Search Answers
   Generates concise answers from indexed content like knowledge articles or documents in response to user queries.
2. Extract Product Mentions
   Identifies and extracts product names or references from unstructured text, such as case descriptions or chat transcripts.
3. Field Generation
   Uses prompt templates to generate content for specific fields on Salesforce records, like call summaries or meeting notes.
4. Flex
   A highly customizable prompt type for general-purpose AI tasks, giving you full control over inputs and outputs.
5. Knowledge Answers
   Retrieves and summarizes information specifically from Knowledge Articles to answer questions accurately.
6. Record Summary
   Creates summaries of Salesforce records, such as Opportunities, Cases, or Contacts, by analyzing record data and related activity.

We will use field generation and flex types for this post.

Real-World Applications

Prompt template is an excellent tool when summarizing the history of a customer and producing recommended actions. This tool can also expedite the process of documenting the actions taken.

Automating the generation of coaching notes can reduce manual processing time from 30 minutes to under 5 minutes.

Data Model

For this use case, I picked a simple data model. The coaching session object is related to the contact for the client. A coaching session record will be created whenever a session is scheduled or delivered. The status field determines whether a session is scheduled or completed. When a new session is scheduled, the date and time of the session is stamped via a record-triggered flow on the client contact record as the next (upcoming) session date and time.

The client contact record also has fields that hold the client’s subscription information and their interests.

Most importantly, all the past coaching session notes need to be considered when preparing for the next session.

Field Generation Prompt Template

The coaching session object includes the preparation notes rich text field. When a new coaching session is scheduled a new record is created. Before the coach goes to this session, they can go into this record and generate preparation notes for their upcoming session.

Field Preparation and Field Generation Prompt Template Creation

Field generation prompt template receives the coaching session record as input. When setting up this prompt template type, you need to scope it for a text or rich text field. Therefore, remember to create this field before you open your prompt builder. In this case, we will scope this prompt template for the preparation notes rich text field.

Record Input

Your prompt template will access all field values of the record it is built on. You can also traverse to related records. You can use field values of the parent records, but referring to related records this way is not very helpful. Your prompt template cannot see a meaningful summary of the related records.

Template-Triggered Prompt Flow Input

This is where flow input comes in handy. You can create a Template-Triggered Prompt Flow and make that flow accessible from your prompt template. Based on my experience, I can say it is quite tricky to make this flow available inside your prompt builder resource picker. I think there is a bug that is scheduled to be fixed soon, that is preventing the flow from showing up in the prompt builder in some cases.

For this use case, I built a flow that retrieves all the past completed coaching sessions and pastes the relevant information into the prompt for the template to consume.

Building the flow is not that difficult, but you will need to loop if you want to summarize multiple records.

Here is how you build the flow:

• Start a Template-Triggered Prompt Flow.
• Configure the start element: Choose Automatic Inputs, Field Generation Template Capability Type and point your start element to the Coaching Session object.

• Get all completed coaching sessions related to the same contact on the triggering record.
• Loop the coaching sessions.
• Add an Add to Prompt element inside the loop, and include the relevant field values in the prompt. Here is what mine looked like:

---Here is one completed past coaching session details: Title: {!Loop_Sessions.Title__c}Preparation Notes: {!Loop_Sessions.Preparation_Notes__c}Session Notes: {!Loop_Sessions.Session_Notes__c}Duration (in hrs): {!Loop_Sessions.Duration__c}Status: {!Loop_Sessions.Status__c}---

• • Add another Add to Prompt element outside the loop before the end of your flow and add the following text. This is important for when no prior completed records can be found.

There are no other completed coaching sessions found for this client.

• • Save, debug and activate your flow.

This is what your flow should look like.

Writing the Prompt Template Instructions

Important things to watch for:

• • • • There are effective reusable instructions you will want to save for the future. Therefore, create a text document and save general instructions and guard rails into this document. You can paste these into your future templates later.
      • Start with, “Your job is,” and give the LLM a job description.
      • Continue with context to ground. Drop in relevant field values here. Remember to pick your field values using the resource picker.
      • Then give the template a summary of the past history from the flow. You should be able to see your flow in your resource picker. If you cannot, you may need to save, deactivate, reactivate, try a few things. I had difficulty getting this done to be honest.
      • The large language model does not understand checkbox values or picklists automatically. You need to define and explain what these values mean, and what actions need to be taken.
      • As you write these, be sure to set clear guidelines and include both what to do and what to avoid. Additionally, specify how the response should be structured to keep everything consistent and easy to follow. For example:
        • • We are writing for a Salesforce Rich Text Field, so you’ll need to follow this guidance below:
            • • • Provide output in HTML format. Only Output the text inside the tag pair.
                • Use bullets or numbered lists to highlight key points.
                • Enhance readability by adding rich text formatting (i.e., font, text color, background color, etc.) where possible.
                • Add HTML tags to improve readability.

Save your prompt template and test it. Save a new version when needed. You will need to point the prompt template to an existing record in Salesforce, therefore create a coaching session record before you test. Read the resolution and response carefully. Make iterations and improve the template. Save and activate your work.

Field Generation Prompt Template Content

Here is my complete prompt template content. You can copy and paste this to get started. Remember to delete the inputs, and insert them from your resource picker for accuracy.

Your Job is to provide a note and guidance for the Coach on what they need to know for an upcoming Coaching Session. This response will be used to populate the session preparation notes field on the Coaching Session record.Use the following context to ground your final responseContext:Client's Name: {!$Input:Coaching_Session__c.Contact__r.Name}Client's Birthday: {!$Input:Coaching_Session__c.Contact__r.Birthdate}Client's Next Scheduled Session Date: {!$Input:Coaching_Session__c.Contact__r.Next_Scheduled_Session__c}Is Client a Subscriber: They are if the value of  is true.Client's Subscriber Type: {!$Input:Coaching_Session__c.Contact__r.Subscriber_Type__c}Client's Interest Areas: {!$Input:Coaching_Session__c.Contact__r.Interests__c}Recent coaching sessions (these are the recently completed coaching session notes separated by "---", extract past experience and interests): {!$Flow:Coaching_Session_PTF_Past_Completed_PT_Flow.Prompt}Note on Subscriber Types:There are three types: "Career", "Personal" and "General"- "Career" focuses on progress and advancement at the work place- "Personal" focuses on wellbeing and happiness outside of work.- "General" is a healthy mix of the two other types.Note on Session Types:There are 2 types of Session Types Inaugural (first) and Follow-up sessions. Specify which type is relevant and the duration of the session. If there are one or more than one completed past coaching session records, then the Follow-up session type will be used, otherwise Inaugurual will be set up.- Inaugural: If the client does not have any previous consultation on record, then this format will be used. This coaching session will be guided toward setting up a plan with the client. Discussions should be centered around making sure they have the proper readiness to start, discussing their interests, discussing their past experience, providing them with a few guidance tips on getting started, and telling them what to expect. Usually go for 1.5 hour. A follow-up is usually scheduled for one month after the first one (this session).- Follow-up: These sessions are monthly check-ins with the customers to review their progress and answer questions for them based on what they are dealing with. Usually go for 1 hour.Note on birthdays: Customers with upcoming and past birthdays, within 30 days of the next session scheduled date ( {!$Input:Coaching_Session__c.Contact__r.Next_Scheduled_Session__c}) will get a free digital book. Please add this detail into the notes if the birthdate is within 30 days of the next scheduled session. If no birthday is provided on the client contact record, or the birthdate is not within 30 days of the next scheduled session do not include this information.- If {!$Input:Coaching_Session__c.Contact__r.Is_Subscriber__c} value is false then the client is not subscribed. Warn the coach: Include a message saying confirm subscription before visit. If the client is subscribed (value = true) do not include a message for the coach about this.Response:- Your response should give the coach information to go into the consultation and know what to discuss with the customer.- Your response should contain important information about the customer, their interests, structure on how the coaching session should go, guidance for their client based on interests and past experience.- Use break-lines in your response to segment between different ideas- We are writing for a Salesforce Rich Text Field, so you'll need to follow this guidance below:- Provide output in HTML format. Only Output the text inside the tag pair.- Use bullets or numbered lists to highlight key points.- Enhance readability by adding rich text formatting (i.e., font, text color, background color, etc.) where possible.- Add HTML tags to improve readability

Adding the Prompt Template to the Page Layout

Once the prompt template is finished, you need to add the prompt template to the page layout. Go to the object lightning page layout, activate dynamic forms if not already active, pick the preparation notes field and link it to the prompt template you have created.

Field Generation Result

Here is the result of the field generation prompt template on the record.

Flex Prompt Template

Flex prompt template is a more powerful tool. You can set it up using the prompt text you have already created. Flex prompt template can accept up to five different inputs of various different types. They can be built into Agentforce actions and used by Agentforce topics.

Currently, prompt template inputs cannot be changed once configured. You will need to create a new prompt template if you want to change inputs. Therefore, always add at least one free text input when configuring a flex prompt template. This will give you the flexibility you need, if you need to add additional instructions or data to your prompt.

Object Preparation and Flex Prompt Template Creation

The input types you can use, when creating a flex prompt template are:

• Object
• Free Text
• Data Model Object

For this template type, you can input any object you want. They don’t have to be related. You need to makes sure you create you object before you build your template, if you are working with custom objects.

You also don’t specify fields when you build a flex template. Your template will have access to all the fields on the record that will be inputted into the template.

Flex Template Inputs

For this use case we will use three input variables:

• Coaching Session Id
• Client Contact Id
• Free Text Input

We could use just the coaching session Id here, but the additional inputs will give us flexibility in how we process different scenarios. For example: We may decide to use this prompt template even before the first session is scheduled, therefore prompt template could work just by receiving the client contact Id.

Reuse Template Triggered Flow Input

I found that I can reuse the same flow I created for the field generation prompt template. This flow should be accessible for you in your resource picker, when you are building your prompt.

Writing Flex Prompt Template Instructions

The prompt template content for this will be very similar to the field generation example. The main difference will be that you will refer to three inputs and the flow input for this template type. You can start with your own template from the previous example and modify it to fit the needs of the flex template. Or you can use my content provided in the code block below to get started.

Once you finish your first pass, save your prompt template and test it. Save a new version when needed. You will need to point the prompt template to existing records in Salesforce, therefore create a coaching session record before you test. You can add additional data and instructions in the free text input field. Read the resolution and response carefully. Make iterations and improve the template. Save and activate your work.

Flex Prompt Template Content

Here is my complete prompt template content. You can copy and paste this to get started. Remember to delete the inputs, and insert them from your resource picker for accuracy.

Your Job is to provide a note and guidance for the Coach on what they need to know for an upcoming Coaching Session.Use the following context to ground your final responseContext:Client's Name: {!$Input:ClientIdVar.Name}Client's Birthday: {!$Input:ClientIdVar.Birthdate}Client's Next Scheduled Session Date: {!$Input:ClientIdVar.Next_Scheduled_Session__c}Is Client a Subscriber: They are if the value of {!$Input:ClientIdVar.Is_Subscriber__c} is true.Client's Subscriber Type: {!$Input:ClientIdVar.Subscriber_Type__c}Client's Interest Areas: {!$Input:ClientIdVar.Interests__c}Recent coaching sessions (these are the recent completed coaching sessions that will give you interest and past experience information):{!$Flow:Coaching_Session_PTF_Past_Completed_PT_Flow.Prompt}Note on Subscriber Types:There are three types: "Career", "Personal" and "General"- "Career" focuses on progress and advancement at the work place- "Personal" focuses on wellbeing and happiness outside of work.- "General" is a healthy mix of the two other types.Note on Session Types:There are 2 types of Session Types Inaugural (first) and Follow-up sessions. If there are one or more than one completed past coaching session records, then the Follow-up session type will be used, otherwise Inaugural will be set up.- Inaugural: If the client does not have any previous consultation on record, then this format will be used. This coaching session will be guided toward setting up a plan with the client. Discussions should be centered around making sure they have the proper readiness to start, discussing their interests, discussing their past experience, providing them with a few guidance tips on getting started, and telling them what to expect. Usually go for 1.5 hour. A follow-up is usually scheduled for one month after the first one (this session).- Follow-up: These sessions are monthly check-ins with the customers to review their progress and answer questions for them based on what they are dealing with. Usually go for 1 hour.Note on birthdays: Customers with upcoming and past birthdays, within 30 days of the next session scheduled date ( {!$Input:ClientIdVar.Next_Scheduled_Session__c} ) will get a free digital book. Please add this detail into the notes if the birthdate is within 30 days of the next scheduled session. If no birthday is provided on the client contact record, or the birthdate is not within 30 days of the next scheduled session do not include this information.Note on Recent Coaching Session records: The recent orders should contain client's interests and past experience.If there are additional notes, they will be provided in {!$Input:FreeTextVar}. Please take these into consideration.Response:- If {!$Input:ClientIdVar.Is_Subscriber__c} value is false then the client is not subscribed. Warn the coach: Include a message saying confirm subscription before visit. If the client is subscribed (value = true) do not include a message for the coach about this.- Your response should give the coach information to go into the consultation and know what to discuss with the customer.- Your response should contain important information about the customer, their interests, structure on how the coaching session should go, guidance for their client based on interests and past experience.- Use break-lines in your response to segment between different ideas- We are writing for a Salesforce Rich Text Field, so you'll need to follow this guidance below:- Provide output in HTML format. Only Output the text inside the tag pair.- Use bullets or numbered lists to highlight key points.- Enhance readability by adding rich text formatting (i.e., font, text color, background color, etc.) where possible.- Add HTML tags to improve readability

Adding Your Flex Prompt Template to Agentforce

I will cover this topic in an upcoming post.

Conclusion

As artificial intelligence continues to redefine the landscape of business processes, tools like Salesforce’s Prompt Builder stand at the forefront of this technological revolution. The insights provided by Jaswinder Rattanpal at TrailblazerDX 2025 inspired this demo and the post. There is no doubt everyone needs to master prompt engineering to prepare for the AI future.

For Salesforce users and enthusiasts looking to discover more about the capabilities of Prompt Builder, the platform provides comprehensive learning and support through Trailhead, hands-on workshops, and a vibrant community forum. Finally, remember to check back at Salesforce Break for additional posts on Prompt Templates and Agentforce AI.

Explore related content:

Salesforce AI: Transforming Data into Sales Engagement

New Agentforce Specialist Certification

How to Get Your AI-Powered Enhanced Developer Org with Agentforce and Data Cloud

#Agentforce #AI #AISpecialist #Einstein #PromptBuilder #Salesforce #SalesforceAdmins #SalesforceDevelopers

Stop Agentforce Dev Orgs From Expiring

Have you ever discovered—often too late—that the Salesforce Dev Org you worked so hard to set up has suddenly expired? You spend hours or even weeks configuring a demo, exploring new features, or practicing your development skills, only to log in one day and be greeted by an expiration notice. It happens more frequently than we realize, especially given that new specialty and trial orgs sometimes have shorter lifespans. A one-month trial here, a week-long environment there—before you know it, all that effort can vanish without warning.

Dev Orgs Don’t Last As Long As They Did In The Past

In the past, standard Salesforce developer orgs might have lasted six months or even a year. Then came specialized developer orgs for different industries, along with new trial orgs that offered limited time frames. Eventually, Data Cloud and now Agentforce Dev Orgs arrived. While these are fantastic for exploring cutting-edge features—like real-time data capabilities and new AI solutions—they usually come with a much shorter shelf life. Salesforce has recently announced new Agentforce Data Cloud developer orgs that expire after 45 days. It is indeed better than just a week, but the clock is still ticking.

Salesforce does send notification emails to remind users that their developer orgs are nearing expiration. However, these alerts can sometimes blend in with other automated messages, or they may land in spam folders. Whether you are a seasoned developer juggling multiple environments or a newcomer trying to learn the platform, it can be easy to lose track of which org is about to expire and which is still active.

The Challenge: Tracking Expiration Before It’s Too Late

Consider how critical these development and testing environments can be. They allow you to:

• Build proof-of-concept solutions without risking a live production org.
• Experiment with new Salesforce features or packages in a controlled setting.
• Demo functionality to colleagues, stakeholders, or potential clients.

If one of these invaluable environments suddenly expires, you may lose crucial configurations, data setups, and more. The result is wasted time and frustration as you scramble to recreate everything in a fresh org.

While the standard practice is to keep track of all login details in a password manager or a spreadsheet, not everyone is consistent with that. Even if you meticulously store your credentials, you might not remember to log in regularly to reset the expiration countdown. That is exactly why an automated reminder can save you from this common pitfall.

Introducing a Schedule-Triggered Flow to Send Automatic Warnings

The solution is surprisingly straightforward: create a scheduled trigger flow in Salesforce that checks for inactivity in your Agentforce or Data Cloud developer org. If the system sees that you have not logged in for a certain number of days (in this case, 40 days), it sends you an email alert. This way, you will have time to log in and reset the expiration clock before day 45 sneaks up on you.

You can build this flow yourself fairly easily, or you can install an unmanaged package I have created. This package contains a flow that runs each night to review login history. If the flow sees that your last login is older than the set threshold, it pings your email to remind you to hop back in. This single step could spare you hours of rebuilding or, worse, losing all the work you have done.

Below is a walkthrough of how to set up this scheduled flow. You will also find tips for customizing it to your specific username if you have changed it from the default (which normally ends with @agentforce.com).

Step 1: Confirm Your New Developer Org’s Basics
First, head to the Setup menu. Check your username and confirm that it ends with @agentforce.com. Confirm that your email address is accurate. You can also check that the Org URL includes “org farm,” indicative of the new Agentforce and Data Cloud Dev Orgs.

Agentforce and Data Cloud developer orgs often come with constraints—particularly around how many AI requests you can make. That said, for most learning or exploration scenarios, these limits will not be prohibitive.

Step 2: Build or Install the Flow
If you opt to build the flow from scratch, you will create a Scheduled Trigger Flow that runs once daily. Set it to run near midnight to ensure it checks your org’s login activity for that day.

In this flow, you will do the following:

1. Define a constant (e.g., usernameSearchString) that points to any user whose username ends with @agentforce.com.
2. Query the User object to find any records whose Username matches that constant. In most new developer orgs, there should only be one match.
3. Retrieve the LoginHistory for that user, sorting records in descending order by login time. This step ensures you get the most recent login date.
4. Compare that login date to the current date minus 40 days. If the last login is older than 40 days, proceed to send an email alert. If it is not, the flow does nothing.

If you have changed your default username, you will need to adjust your constant to match the portion that identifies your user. Otherwise, the out-of-the-box approach (i.e., checking for @agentforce.com) will work fine.

Step 3: Configure the Email Alert
Within the flow, you will set an Action to send an email. This email typically goes to the address tied to the user record (the one you provided when creating the developer org). The email is straightforward but crucial. You might specify a subject line like “Warning: Your Developer Org Will Expire Soon” or something equally attention-grabbing.

For the email body, you can create a Text Template. Include pertinent details such as the last login date, the username, and a short explanation that logging in again will reset the expiration countdown. Include enough information that you can quickly locate the correct credentials in your password manager. A typical text template might read:

Body:
Hello, This is an automated reminder that your Agentforce developer org, associated with username {!User.Username}, is nearing its 45-day expiration window. Your last login was {!LoginHistory.LoginTime}. To prevent expiration, please log in as soon as possible.

Thank you!

You can, of course, modify the text to your preference. The main goal is to ensure you see this email and know exactly which org requires attention.

Step 4: Activate and Confirm the Flow
When you install the flow template from the unmanaged package, it might come pre-activated. If you build one from scratch, remember to Activate it. Then, check under Scheduled Jobs in Setup to confirm the flow is listed to run nightly.

If everything is configured correctly, you will receive an email once your last login date surpasses 40 days. At that point, you know you have a five-day grace period before the 45-day expiration hits—plenty of time to jump back into the org and keep it alive.

Step 5: Enjoy Peace of Mind
That is it! No more frantic searches through your password vault only to discover your demo is gone. The next time you spin up a new developer org for Data Cloud exploration, build this simple safety net. You will save yourself the frustration of expired orgs and lost work.

Install the Unmanaged Package to Leverage the Scheduled Flow

The link for the unmanaged package is as follows:

https://login.salesforce.com/packaging/installPackage.apexp?p0=04tgK0000000TOb

Bonus: How To Add the Org URL to the Email

One of our YouTube subscribers asked whether the Org URL can be added to the email. This is a very good suggestion, and I am planning on adding it to the unmanaged package in the future.

If you want to add the URL, please follow these instructions:

1. Create a URLFormula resource. Your formula will read as follows:LEFT($Api.Partner_Server_URL_630, FIND( '/services', $Api.Partner_Server_URL_630))
2. Go to your email body Text Template and view source (view as Plain Text), and replace the username resource {!Get_User.Username} with the following:
   <a href="{!URLFormula}">{!Get_User.Username}</a>

This will make the username value on the email clickable. Once the user clicks on the username, they will be taken to the login page for the Org.

Final Thoughts on Agentforce Data Cloud Developer Orgs

These new developer environments are wonderful for trying out the latest Salesforce innovations. You can test specialized features that may not be available in a more traditional developer edition. The big trade-off is their comparatively short expiration period. Forty-five days is enough for many use cases, yet it is still short enough that forgetting just one login can lead to losing your entire environment.

With the scheduled trigger flow approach, you can have the best of both worlds: a flexible, feature-rich developer org with a relatively short lifespan, backed by a reliable reminder system preventing accidental expirations.

Watch the video on YouTube for additional details 👇

Explore related content:

Salesforce Summer ’25 Preview: Major Flow Changes to Watch For

How to Get Your AI-Powered Enhanced Developer Org with Agentforce and Data Cloud

TDX 2025 News: Salesforce Agentforce 2dx and AgentExchange

My Inaugural Agentforce-Assisted Flow Building Experience

#Agentforce #DataCloud #DevOrg #Salesforce #SalesforceAdmins #SalesforceDevelopers

Agentforce Hackathon TDX25: My Hackathon Tips

When Salesforce opened registration for TDX 2025 (TrailblazerDX), they also announced the Agentforce Hackathon, which was to be held before the main event. The event was free to attend and it came with a bunch of prizes including a grand prize of $100,000.

Hackathon

A hackathon is a fast-paced event, usually lasting from one to several days, where people collaborate intensively on technology projects. Participants—such as developers, designers, and product managers—team up to tackle challenges or create new solutions from scratch.

Key features of a hackathon include:

1. Collaboration: People from different backgrounds work in small teams to develop a proof-of-concept, prototype, or even a fully functioning product.

2. Innovation and Creativity: There’s usually a focus on rapid ideation and experimentation. This environment encourages participants to test out new ideas or use emerging technologies.

3. Time Constraint: A hackathon often runs over an intense, short period—like 24 or 48 hours—though some may last for several days. The time limit pushes teams to produce results quickly.

4. Competition or Challenge: Many hackathons incorporate prizes or judging criteria around best idea, best design, or most innovative project. This gamified element helps motivate participants.

5. Community and Learning: Hackathons also act as social and educational events. Newcomers can learn from more experienced participants and meet potential collaborators or employers.

Hackathons are hosted by a variety of organizations, from tech companies and startups to universities and nonprofits, each often targeting a specific theme (e.g., healthcare, finance, machine learning, social impact). They can serve as a way to spur creativity, recruit talent, or generate new ideas and products in a short period of time.

Salesforce Agentforce

Salesforce Agentforce is a digital platform that enables organizations to build and deploy autonomous AI agents across various business functions. These agents can perform tasks such as customer service, sales development, and marketing automation, operating 24/7 to enhance productivity and efficiency. Agentforce integrates seamlessly with existing Salesforce applications, allowing businesses to customize agents to their specific needs, thereby augmenting their workforce with AI-driven capabilities.

Agentforce Hackathon

For the Agentforce Hackathon you were asked put together a team of max 5 people. You could build your team before getting to the Hackathon, or pick your team there. Although initially it looked like all TDX participants could join, Salesforce later announced what type of participants will be allowed. International participants from many countries, recent Salesforce employees and Salesforce vendors were ineligible to participate in the event due to legal constraints.

The event started at 4:00 PM on Monday and continued until midnight that day. The teams came back the next day as early as 7:00 AM and worked until latest at 8:00 PM on Tuesday. Food, coffee, soda and snacks were provided. Everyone who submitted a survey at the end of the event received a t-shirt. There was a grand prize of $100,000 as well as more vertically focused smaller prizes like best use of Slack. The winners were announced on Wednesday during the keynote. Congrats to the winners.

My Hackathon Experience

Overall the event was simultaneously fun and exhausting. Most people arrived in San Francisco after a long travel, and they were also impacted by the timezone changes.

I did not have a team or an idea ready when I went there. I collaborated with Lynda Kane and Andrea Stratton, two people I know in the Salesforce ecosystem. Collaborating with different brilliant minds was one of the best parts of the experience.

The event required demo skills as you had to record and submit a 5 min long video along with the documentation at the end of the event.

I’m a builder at heart, and I was proud that we had a working solution by the end. That was a key goal for me. Additionally, the event gave me an opportunity to lock in and experiment with Salesforce Agentforce for 13-14 hours. It is a great way to learn the features hands on.

One of my teammates, Lynda, said, “Working with Andy and Andrea for the Hackathon was fun and a great collaborative experience. This was an excellent way to get hands-on with the tools and build on our skills. The way I need to write instructions for prompts and agents is finally starting to click for me.”

Collaboration was a key part of the hackathon, and Andrea Stratton, one of my teammates, found the hands-on challenge to be a valuable learning opportunity: “I appreciated getting to learn from other community members with different skillsets and backgrounds,” she said. “Having to work with Agentforce without a guided walkthrough that we get with Trailhead was a worthwhile challenge and learning experience.”

The Result

Our team successfully completed the Salesforce Agentforce TDX Hackathon. As Team Kudosforce, our solution looks at reviews submitted by customers and creates immediate feedback and actionable insights into Slack to expedite continuous improvement.

It was wonderful to work with Lynda Kane and Andrea Stratton at this event.

Watch the demo video below.

Tips for Hackathons

1. Since the time is very restricted, it helps to build a team before the event and align on working principles.

2. The idea is the most important part, so it would be helpful to discuss it before the event.

3. While it’s important to have an idea you like before Hackathon, you can not produce work and build the solution before the event. This is against the rules.

4. Review the submissions of the previous winners. This is public information. See what resonates with the judges.

5. I’d say a big visionary idea is more important than a practical use case that you can build 100% within the time provided. Some aspects of your solution can be mocked and you can communicate that. That is totally OK.

Conclusion

Participating in the Agentforce Hackathon at TDX 2025 was an unforgettable experience—intense, rewarding, and a fantastic way to push creative and technical limits. The event reinforced the power of collaboration, rapid ideation, and hands-on experimentation with new Salesforce technologies. Whether you’re a seasoned builder or a first-time hackathon participant, the key is to embrace the challenge, stay adaptable, and focus on crafting a compelling solution.

Now that we have Agentforce & Data Cloud dev orgs available, preparing for a hackathon has become much easier. If you’re considering joining a future event, check out the upcoming Agentforce Virtual Hackathon and get involved! With $140,000 in prizes, there are plenty of opportunities to win:

🏆 Overall Grand Prize – $50,000 in cash (1 winner)
🎖 Honorable Mention: Most Impactful – $20,000 in cash (1 winner)
💡 Best Use of Slack Prize – $25,000 in cash (1 winner)
🎨 Honorable Mention: Most Creative – $20,000 in cash (1 winner)
🎤 Honorable Mention: Best Demo Delivery – $20,000 in cash (1 winner)
🗳 People’s Choice (Public Voting) – $5,000 in cash (1 winner)

If you’re up for the challenge, start preparing now! Have questions or insights? Drop them in the comments!

Explore related content:

TDX 2025 News: Salesforce Agentforce 2dx and AgentExchange

How to Get Your AI-Powered Enhanced Developer Org with Agentforce and Data Cloud

New Agentforce Specialist Certification

Your Ultimate Guide to Dreamforce 2024

#Agentforce #Hackathon #Salesforce #SalesforceAdmin #SalesforceDeveloper #TDX25 #TrailblazerDX

🤖 AI
🔴 Salesforce to Invest $1B in Singapore for AI

🔸 Salesforce to boost AI adoption with $1B investment over 5 years.
🔸 Plans include expanding Agentforce and enhancing Singapore Airlines’ AI systems.
🔸 Follows $500M investments in Saudi Arabia & Argentina.

#Salesforce #AI #Agentforce #Singapore #CloudTech

Top 7 Key Takeaways from Salesforce Dreamforce 2025

Salesforce Break reviewed the press releases and sessions coming out of Salesforce Dreamforce 2025, and prepared the ket takeaways in this post, so you don’t have to go thorough all the materials.

The biggest announcements for Salesforce at Dreamforce 2025 were centered on advancing the company’s foundational “Agentic Enterprise” vision through enhanced control, deeper context integration, and widespread collaboration tools.

The announced functionalities were more evolutionary than revolutionary.

Here is the list.

Top 7 Key Dreamforce Takeaways

1. Agentforce 360 Platform, New Agentforce Builder, and Agent Script

The cornerstone announcement was the launch of Agentforce 360, the latest version of the comprehensive platform designed to unify AI, trust, and data capabilities across all Salesforce products. Salesforce has completely reimagined the entire Customer 360 platform as Agentforce 360, ensuring that every app is now “agentic”. This platform emphasizes providing users with more control than ever over their AI systems. To make development accessible to a wider audience, including line-of-business leaders, IT teams, service, and sales teams, a brand new AgentForce Builder was introduced, featuring a radically simplified, clean, and beautiful interface built from the ground up.

This capability is powered by Agent Script, a new scripting language that exposes the reasoning engine and allows users to define deterministic chaining actions and conditional logic. Agent Script unlocks patterns needed for mission-critical use cases, blending fluid agentic reasoning with the certainty of rules-based control in a unified instruction set to ensure agents are predictable and stay “on track,” preventing costly unpredictability. Agent Script can be built at the topic level where previous LLM-based non-deterministic functionality produced unpredictable results.

In addition, Salesforce announced Slack as its future conversational interface. Several sessions demonstrated deeper integrations in action. Another major change of course was the ability to use external LLMs for the Atlas Reasoning Engine. I believe this demonstrates that Salesforce is positioning Agentforce more as an orchestrator and collaborator of agents and AI capabilities rather than competing to become the agent for the enterprise.

2. Agentforce Voice

Agentforce Voice extends the power of the Agentforce platform by allowing agents to talk, bringing AI capabilities directly to contact centers and 800 numbers. Businesses can now configure the voice, tone, and personality of the AI right inside the AgentForce Builder. The goal is to deliver a unified customer experience across all channels, providing a highly human-like and interruptible conversational flow. A critical feature of Agentforce Voice is ensuring a seamless transition when the AI needs to transfer a customer to a human agent; the human representative automatically receives the full transcript and context of the AI conversation, allowing them to pick up the experience precisely where the AI left off. This functionality is available GA October ’25.

3. Intelligent Context Processing (Data 360)

Intelligent Context Processing tackles one of the greatest challenges for AI agents: understanding and utilizing vast amounts of complex, unstructured data. Agents often struggle with content in rich formats, pictures, tables, and existing workflows, the accumulated wisdom of the company. These new tools, built into Data 360, interpret and index this data by analyzing and parsing complex content (such as product manuals containing charts and images). This allows agents to pull in the exact, correct context required to deliver accurate and rich responses at the precise moment it is needed.

Furthermore, Data 360 enhances governance across both structured and unstructured data. Using natural language, administrators can create policies, such as masking internal FedEx employee contact details within agent responses, ensuring the information provided is not only accurate but also appropriate for the customer. It is not clear to us whether this is solely a rename of the product called Data Cloud. It seems that way.

4. Agentforce Vibes

Salesforce launched Agentforce Vibes as a new product that lets trailblazers quickly and easily build apps, dashboards, automations, and flows. Users achieve this via vie coding, which involves providing a simple, natural language description of what they want the platform to build. The core innovation of Agentforce Vibes is its deep contextual understanding; it speaks “the language of the business,” including the organization’s data, relationships, customers, products, and security permissions. This contextual intelligence allows Agentforce Vibes to rapidly translate a descriptive idea into deployable, production-grade Salesforce metadata (such as a screen flow). This drastically reduces development time, saving what could amount to dozens of manual clicks inside a traditional flow builder. This effectively elevates the capabilities of every developer. Interesting tidbits: Developers can develop using the coding language of their choice, and there is a local LWC preview function that will be launched soon.

5. Slackbot

Salesforce unveiled Slackbot as a new personalized, AI-powered companion that boosts productivity directly within Slack. It will launch for General Availability (GA) in January and draws on each user’s unique organizational context, including conversations, files, workflows, and apps. The tool moves users beyond asking simple questions toward achieving complex, tangible outcomes. For example, a user can ask Slackbot to handle a multi-step process with one command. It can review deal status, find compliance files, and draft a customer email in the user’s tone. Slackbot can also create a prep document and calendar invite for key stakeholders automatically. Slackbot will be the home of AI capabilities within Slack, even for customers who don’t use Salesforce.

6. Support for Third-Party Agents in Slack (AgentExchange)

Salesforce affirmed its vision of Slack becoming the “home for all AI” by announcing support for all third-party AI agents, such as ChatGPT, Claude and Gemini. This transformation positions Slack as an agentic operating system where external agents can exist as collaborative “teammates” alongside human employees. To ensure these external agents can perform sophisticated reasoning, they are grounded in the company’s real-time knowledge and context via a real-time search API and an MCP server. This initiative allows Salesforce agents to work in conjunction with agents from other platforms. This coupled with the AI-assisted enterprise search capabilities of Slack empowers Slack admins and users to be more productive.

7. Agentforce Observability

Agentforce Observability was introduced to help monitor and scale digital work in the new agentic enterprise era. It serves as one control center for managers to monitor and improve agent team performance. The tool gives leaders visibility into KPIs like escalation and deflection rates using Tableau Next analytics.

Most importantly, it features Agent Insights, which acts as a performance review by scoring every single agent session. This scoring helps managers find and analyze poor-performing conversations to uncover root causes like process issues. It enables tuning of agent prompts and behaviors for consistent results. This management layer is essential since prompts and loops alone aren’t enough.

This was a major pain point with the clients. I am happy Salesforce is addressing it with this new functionality that will be available for most clients.

Conclusion

I personally found the announcements more evolutionary than revolutionary. It was not a strong Dreamforce in terms of new functionalities covered.

Adoption challenges and cleanup are still needed to make current products appealing. These announcements mark real progress for Salesforce.

[youtube https://www.youtube.com/watch?v=_S18LAXcBYQ?version=3&rel=1&showsearch=0&showinfo=1&iv_load_policy=1&fs=1&hl=en-US&autohide=2&wmode=transparent&w=640&h=360]

Explore related content:

Salesforce Ushers in the Age of the Agentic Enterprise at Dreamforce 2025

Dreamforce 2025: Standout Sessions Streaming on Salesforce+

Salesforce Winter ’26 Release: Comprehensive Overview of New Flow Features

Dreamforce is now in the rearview mirror, so I’d like to give a few of my own reflections on the event. The one term that took center stage in Salesforce’s ecosystem: Agentforce. Introduced only weeks ago, Agentforce is now integrated across multiple platforms like Slack and Tableau, positioning itself as a key player in Salesforce’s vision for the future. This article explores how Agentforce, Data Cloud, and new trends like consumption-based pricing are reshaping Salesforce, and what it all means for the future of CRM.

Agentforce is everything

In a meeting with Salesforce right before Dreamforce, I heard Salesforce experts answer the ultimate question about Agentforce: What is Agentforce? Is it the new Copilot or Einstein bot? The answer was “Everything is an agent.” When I heard that, I should have known what was coming in Dreamforce. Almost every session had Agentforce in it. We just heard the term for the first time a few weeks back, and Salesforce said Agentforce is GA (generally available) in Slack, GA in Tableau, etc. That’s because all existing Copilot solutions got relabeled to Agentforce. What sets Agentforce apart from Salesforce’s previous pitches in this area?  Salesforce emphasizes that Agentforce can take actions. And Salesforce will provide you with a bunch of out of the box actions. If you want to learn more about Agentforce head over to our dedicated blog post HERE.

Data Cloud is still important

When I reviewed the schedule for Dreamforce, I saw that Data Cloud still had a keynote, but I was skeptical. Will we forget about Data Cloud’s notoriety in a few years? I am still not sure. However, after watching the sessions, I see that Data Cloud is required to deploy Agentforce. Whether that will remain so in the future, I am uncertain. I can tell you for now that Data Cloud is here and still holds a good share of the center stage.

Consumption-based pricing is the way forward

There is a general trend among the clients: They are tired of fixed per-seat subscription pricing. This is not isolated to Salesforce, by the way. I heard that the list price for Agentforce is $2 per interaction. Most of you know that Data Cloud is also sold based on consumption-based pricing models. There is a general trend in Salesforce’s approach, which is expected to continue: We are heading to a consumption-based pricing future.

Conversational Interface is where it’s at

When we watched these Dreamforce sessions, many of us did not realize that Salesforce is setting a vision for how users are expected to interact with their CRM in the future. The future is conversational. Clicking buttons, turning sliders on and off, or dragging items from one box to another will be a thing of the past. Instead, you’ll converse with your CRM. Typing may be necessary for now, but soon, you’ll be able to talk and listen.

ROI takes center stage instead of Trust

When AI came along, Salesforce sold very hard: It is the only provider that can provide the secure Trust layer you need. Salesforce was the Sheriff in the Wild West. After more than a year, it is evident that the Trust pitch did not work. Do we not have the Trust Layer anymore? We do. But the focus has changed. All demos focus on the ROI of the implementation rather than the Trust Layer aspect.

Slack’s Salesforce Channels: The first attempt to dethrone Chatter

Let’s face it: Until now, Slack has not done that much to integrate with Salesforce. AI capabilities have also been somehow slow as well. At Dreamforce, Salesforce announced Salesforce Channels in Slack. Now you can create a channel in Slack for your Salesforce object records. Your record field values now appear in Slack, and the Slack conversations are displayed on the Salesforce page layout. Sounds very promising, does it not? We will see how well it is implemented when it is GA. For the Slack and Salesforce clients this could be the start of the end of Chatter.

Salesforce Foundations comes bearing gifts

Salesforce announced Foundations, a free upgrade that brings key cross-department features to every Salesforce customer at no additional cost. Foundations includes Sales, Service, Marketing, Commerce, and Data Cloud features, all in a new, enhanced UI. This makes it easier for businesses to get a unified view of their customers and build connected, AI-powered experiences across every customer touchpoint. Salesforce has been making worthwhile attempts to bring various cloud capabilities into core under the same umbrella making it easy for smaller clients to adopt new features.

Admin Role is changing

I watched the admin keynote, and the content did not seem to be geared toward admins. What is the reason for that? There are two possibilities:

1. Salesforce is pushing hard to sell Data Cloud and Agentforce, and the content was shaped this way due to that motive.
2. Salesforce sees that the admin role needs to change and, therefore, is pushing the discussion in this direction.

In the admin demos, we saw user management, Data Cloud, Agentforce, integrations, flow, and invocables. The content was a mashup of pretty advanced capabilities. Are the admins expected to upskill to this scope? Will the new admin job posts include all these capabilities? I don’t know. What I know for sure is that the change is coming. Prepare for the inevitable by getting AI certified for FREE for a limited time.

What do you think?

Join the conversation.

Explore related content:

Top 9 Salesforce Winter 25 Flow Features

What is the Salesforce AI Specialist Certification?

Exploring Salesforce Admin Certification Paths

https://salesforcebreak.com/2024/09/24/dreamforce-2024-reflections/

#Admin #Agentforce #AI #ArtificialIntelligence #Certification #DataCloud #Dreamforce #Foundations #ROI #Salesforce #Slack #Update

»#Salesforce launches #Agentforce #2dx, letting #AI run autonomously across #enterprisesystems.« https://venturebeat.com/ai/salesforce-launches-agentforce-2dx-pushing-autonomous-ai-deep-into-enterprise-workflows/?eicker.news #tech #media

Top 7 Salesforce Spring 25 Flow Features

Salesforce Releases are triannual updates (Spring, Summer, and Winter) that introduce new features, enhancements, and fixes to the Salesforce platform. These updates are automatically applied to Salesforce environments on an announced schedule. This ensures all users benefit from the latest innovations without requiring manual upgrades. Salesforce Release the Spring 25 functionality to the Preview Developer Orgs.

Preview Developer Orgs allow users to explore and test new features before the release becomes generally available. Salesforce provides sandbox environments or developer editions during the release preview window with the upcoming release functionality. This enables developers and admins to evaluate changes, test compatibility, and prepare their organizations for the new features, ensuring a smooth transition when the release is applied to production environments.

If you can access a preview Org, you can easily experience these new, user-friendly functionalities:

1. Reactive Screen Actions

Screen Flow Reactivity in Salesforce, a feature that enhances efficiency, refers to the capability of screen flow components to dynamically and automatically update based on changes in other components on the same screen without requiring the user to navigate away or manually refresh. This feature improves the user experience by enabling real-time interactivity and reducing the need for complex workarounds or custom code. For instance, when a user modifies a field or selects an option in one component, other components on the screen, such as picklists, formulas, or visibility rules, can immediately adjust their behavior or display relevant data accordingly. Screen Flow Reactivity enhances efficiency in building and using flows by providing a smoother, more intuitive interaction.

Reactive Data Fetching in Screen Flows

In recent releases, the Salesforce flow product team has been rolling out reactivity functionality for screen flows. One popular request for new functionality has been getting data from the org reactively based on user selections on the screen. Josh Dayment’s Data Fetcher package has filled this gap well until the product team can roll out the functionality.

With Spring 25, you can create an autolaunched flow to fetch data from the database and deploy it reactively on screen without needing an action button.

🚨 Use case 👇🏼

A territory assignment reference table is stored in a custom object. When an Opportunity is created and assigned to an Account, the ZIP code on the Account determines who should follow up on the opportunity.

When the user selects an Opportunity row on this data table screen component, the Account, the Owner, and the Assigned Person are displayed per the assignment reference table.

Please note that this functionality also fills a gap with the current data table functionality. Since the lookup fields can only show Ids, the reactive screen action can reactively fetch and display the record’s name for the selected row.

2. Progress Indicator for Flow Stages

The Stage Resource in Salesforce Screen Flows is a feature that allows developers to organize screens into logical stages within a flow. It provides a visual indicator for users, typically in a progress bar or stage tracker, showing their position within the flow and the remaining steps. This resource enhances user experience by providing clarity and context, especially for multi-screen flows that involve complex processes. To use the Stage Resource, you define stages and assign each screen to a specific stage during the flow design process. This helps guide users through the flow in a structured and organized manner, reducing confusion and improving completion rates.

Until this release, Salesforce did not offer a visual progress indicator out of the box.

With this release, a progress indicator is enabled by default. You can change its location and type under the advanced flow settings. Every screen is assigned a stage resource, which determines what label is displayed on the progress indicator. Sadly, the progress indicator does not offer a navigation functionality based on user clicks.

3. Immediate Input Validation in Screen Flows

Screen Flow Validation in Salesforce ensures that the data entered by users in screen flow components meets specified criteria before they can proceed to the next screen or complete the flow. Validation rules are set up within the flow builder by adding Validation Rules to individual screen components. These rules define conditions that the input must satisfy, such as required fields, specific formats, or value ranges. If the user’s input does not meet the criteria, an error message is displayed near the relevant component, preventing them from continuing until the issue is resolved. This feature enhances data accuracy and consistency by ensuring that only valid and complete information is submitted through the flow.

Up until this release, the validation functionality in flow was not reactive. Users only saw the validation message after clicking the Next button. With this release, the validation functionality has been enhanced to be entirely reactive, showing the validation message to the user immediately, similar to component conditional visibility.

4. Email Action Attachment Support

The Email Action in Salesforce Flow allows you to automate the sending of emails directly within a flow. This action enables you to configure and send emails to recipients such as users, contacts, leads, or custom email addresses, using Salesforce Email Templates or dynamically generated content. You can define parameters like subject, body, recipient(s), and attachments, tailoring the email to specific conditions or user inputs within the flow. The Email Action is commonly used for notifications, confirmations, follow-ups, or updates. This streamlines communication processes and enhances user engagement with automated, timely emails.

Email action now supports file attachments. To attach one or multiple files to your email, do the following:

1. Get your Content Document object record by name, specifying the full title and the file extension.
2. Each content document can have multiple versions in Salesforce. Get the latest Content Version Id for your Content Document(s). If you have multiple attachments, add the Ids to a comma-delimited list using the assignment element.
3. Enter the content version Id(s) into the Attachment ID parameter of the email action.

Unlike inserting images from files into text templates for the email body, you don’t need to make your file public before attaching it to your email. If the user running the flow can access the file, it can be attached to the email.

Read more here: How To Attach Files Using the Flow Email Action in Salesforce

5. Get Specifying The Number of Records (Limit)

The LIMIT operator in Salesforce Object Query Language (SOQL) restricts the number of records a query returns. By specifying a numerical value after the LIMIT keyword, you can control the maximum number of rows fetched from the database. For example, SELECT Name FROM Account LIMIT 10 retrieves only the first 10 accounts that match the query criteria. This operator is commonly combined with sorting (using ORDER BY) to ensure specific records are prioritized.

With Spring ’25, the flow builder receives this functionality. Now, the flow developer can specify how many records they want returned as a result of the get element. Please note that collection sort provided this functionality but required an additional element. When getting large datasets, use this functionality to ensure your get won’t fetch more than 50K records. Doing so would yield an error in Salesforce. You can use a max limit of 2,000.

6. Transform Element Can Now Join Collections

Flow builder can now combine source collections from related flow resources into a target collection. This could be particularly useful when you need to combine data from Salesforce with external systems, such as merging order records to create a more comprehensive view of customer transactions. The result? A single, unified dataset that can be used for further processing.

The transform element will now take two source collections and use join keys to produce an inner join. The resulting collection includes field values from both source collections. This functionality is especially useful, when you have an Apex defined collection variable you need to send the data to as an output. This custom variable structure can not be defined and produced within the flow canvas, so far.

Read more here: Transform Element Now Supports Join Collections

7. Various UI and Accessibility Improvements

New Flow Modal Enhancements

The Flow Type Picker modal in Salesforce provides an interface that guides users in selecting the appropriate flow type when creating a new one. It displays categorized options for different flow types, such as Screen Flow, Record-Triggered Flow, Schedule-Triggered Flow, and Platform Event-Triggered Flow, along with descriptions and icons for each. The modal also briefly explains each flow type’s purpose and use cases. This helps users choose the one that best aligns with their business process.

Salesforce has reorganized the modal design again, this time focusing on simplifying automation creation. The new design benefits Salesforce automation newcomers by reducing confusion and guiding users to the correct automation type and template.

What is more confusing than before is that the new experience is only visible when you create a new automation from the Automation Lightning App. The modal is still the same when you create a new flow from Setup.

Resource Search

Resources in Salesforce Flow are elements used to store, manipulate, and reference data throughout the flow. They serve as building blocks to define variables, constants, formulas, collections, and more, enabling dynamic interactions and complex logic within the flow. Resources can hold values such as user inputs, system data, or calculated results and are reusable across different parts of the flow. Previously, when you typed ahead in the resource picker, Salesforce did not return all matching resources accessible under different headings. Therefore, the developers needed more clicks to reach the resources they sought. The resource search is now more powerful, returning results from all types of resources.

Flow Versions

Flow Versions in Salesforce allow you to manage multiple iterations of a flow over time. A new version is created each time a flow is saved or modified, enabling you to track changes and roll back to previous configurations if needed. Only one flow version can be active at a time, meaning users will interact with the specific version marked as “active.” While this versioning system is helpful for iterative improvements, it has limitations; having too many versions may clutter the flow management interface. With Spring 25, flow versions are accessible via the top bar of the flow canvas. Developers can see all flow versions inside the flow canvas and open another version using this new UI.

Keyboard Shortcuts

Salesforce provides keyboard shortcuts in the Flow Canvas to enhance productivity and navigation for flow builders. Salesforce made improvements to the keyboard shortcuts with the Spring 25 release. I thank Yumi Ibrahimzade for discovering the new enhancements. Common shortcuts include using Ctrl + Z (Windows) or Cmd + Z (Mac) to undo changes and Ctrl + Y or Cmd + Y to redo. You can also use arrow keys to navigate elements, Delete or Backspace to remove selected elements, and Ctrl + S or Cmd + S to save the flow. Shortcuts like Tab and Shift + Tab allow you to quickly navigate through screen components or fields. These shortcuts streamline flow creation, reducing reliance on the mouse and making building and adjusting flows efficiently easier.

Please note that there could be modifications to this list as we progress through the release preview schedule.

Conclusion

As we wrap up our exploration of the Spring ’25 Salesforce release, it’s clear that these updates are set to significantly enhance how Salesforce users interact with the platform. This Salesforce release brings a host of new features aimed at improving functionality and user experience, from real-time validation in screen flows to the introduction of email action attachment support, which simplifies the way emails are handled directly within flows.

What do you think of the new flow enhancements? Which new feature is your new favorite?

Explore related content:

Einstein Generative AI for Public Sector Solutions – New Release

A New Era for Salesforce Exams: Transitioning from Examity to Webassessor

Gear Up for TrailblazerDX ’25: Registration is Open!

Salesforce Flow Best Practices

Salesforce Summer ’25 Preview: Major Flow Changes to Watch For

#Admin #Agentforce #AI #Automation #Career #Community #Data #Developer #Lightning #LowCode #Release #Salesforce #Tips #Trailblazer #Trailhead

Agentforce Hackathon TDX25: My Hackathon Tips

When Salesforce opened registration for TDX 2025 (TrailblazerDX), they also announced the Agentforce Hackathon, which was to be held before the main event. The event was free to attend and it came with a bunch of prizes including a grand prize of $100,000.

Hackathon

A hackathon is a fast-paced event, usually lasting from one to several days, where people collaborate intensively on technology projects. Participants—such as developers, designers, and product managers—team up to tackle challenges or create new solutions from scratch.

Key features of a hackathon include:

1. Collaboration: People from different backgrounds work in small teams to develop a proof-of-concept, prototype, or even a fully functioning product.

2. Innovation and Creativity: There’s usually a focus on rapid ideation and experimentation. This environment encourages participants to test out new ideas or use emerging technologies.

3. Time Constraint: A hackathon often runs over an intense, short period—like 24 or 48 hours—though some may last for several days. The time limit pushes teams to produce results quickly.

4. Competition or Challenge: Many hackathons incorporate prizes or judging criteria around best idea, best design, or most innovative project. This gamified element helps motivate participants.

5. Community and Learning: Hackathons also act as social and educational events. Newcomers can learn from more experienced participants and meet potential collaborators or employers.

Hackathons are hosted by a variety of organizations, from tech companies and startups to universities and nonprofits, each often targeting a specific theme (e.g., healthcare, finance, machine learning, social impact). They can serve as a way to spur creativity, recruit talent, or generate new ideas and products in a short period of time.

Salesforce Agentforce

Salesforce Agentforce is a digital platform that enables organizations to build and deploy autonomous AI agents across various business functions. These agents can perform tasks such as customer service, sales development, and marketing automation, operating 24/7 to enhance productivity and efficiency. Agentforce integrates seamlessly with existing Salesforce applications, allowing businesses to customize agents to their specific needs, thereby augmenting their workforce with AI-driven capabilities.

Agentforce Hackathon

For the Agentforce Hackathon you were asked put together a team of max 5 people. You could build your team before getting to the Hackathon, or pick your team there. Although initially it looked like all TDX participants could join, Salesforce later announced what type of participants will be allowed. International participants from many countries, recent Salesforce employees and Salesforce vendors were ineligible to participate in the event due to legal constraints.

The event started at 4:00 PM on Monday and continued until midnight that day. The teams came back the next day as early as 7:00 AM and worked until latest at 8:00 PM on Tuesday. Food, coffee, soda and snacks were provided. Everyone who submitted a survey at the end of the event received a t-shirt. There was a grand prize of $100,000 as well as more vertically focused smaller prizes like best use of Slack. The winners were announced on Wednesday during the keynote. Congrats to the winners.

My Hackathon Experience

Overall the event was simultaneously fun and exhausting. Most people arrived in San Francisco after a long travel, and they were also impacted by the timezone changes.

I did not have a team or an idea ready when I went there. I collaborated with Lynda Kane and Andrea Stratton, two people I know in the Salesforce ecosystem. Collaborating with different brilliant minds was one of the best parts of the experience.

The event required demo skills as you had to record and submit a 5 min long video along with the documentation at the end of the event.

I’m a builder at heart, and I was proud that we had a working solution by the end. That was a key goal for me. Additionally, the event gave me an opportunity to lock in and experiment with Salesforce Agentforce for 13-14 hours. It is a great way to learn the features hands on.

One of my teammates, Lynda, said, “Working with Andy and Andrea for the Hackathon was fun and a great collaborative experience. This was an excellent way to get hands-on with the tools and build on our skills. The way I need to write instructions for prompts and agents is finally starting to click for me.”

Collaboration was a key part of the hackathon, and Andrea Stratton, one of my teammates, found the hands-on challenge to be a valuable learning opportunity: “I appreciated getting to learn from other community members with different skillsets and backgrounds,” she said. “Having to work with Agentforce without a guided walkthrough that we get with Trailhead was a worthwhile challenge and learning experience.”

The Result

Our team successfully completed the Salesforce Agentforce TDX Hackathon. As Team Kudosforce, our solution looks at reviews submitted by customers and creates immediate feedback and actionable insights into Slack to expedite continuous improvement.

It was wonderful to work with Lynda Kane and Andrea Stratton at this event.

Watch the demo video below.

Tips for Hackathons

1. Since the time is very restricted, it helps to build a team before the event and align on working principles.

2. The idea is the most important part, so it would be helpful to discuss it before the event.

3. While it’s important to have an idea you like before Hackathon, you can not produce work and build the solution before the event. This is against the rules.

4. Review the submissions of the previous winners. This is public information. See what resonates with the judges.

5. I’d say a big visionary idea is more important than a practical use case that you can build 100% within the time provided. Some aspects of your solution can be mocked and you can communicate that. That is totally OK.

Conclusion

Participating in the Agentforce Hackathon at TDX 2025 was an unforgettable experience—intense, rewarding, and a fantastic way to push creative and technical limits. The event reinforced the power of collaboration, rapid ideation, and hands-on experimentation with new Salesforce technologies. Whether you’re a seasoned builder or a first-time hackathon participant, the key is to embrace the challenge, stay adaptable, and focus on crafting a compelling solution.

Now that we have Agentforce & Data Cloud dev orgs available, preparing for a hackathon has become much easier. If you’re considering joining a future event, check out the upcoming Agentforce Virtual Hackathon and get involved! With $140,000 in prizes, there are plenty of opportunities to win:

🏆 Overall Grand Prize – $50,000 in cash (1 winner)
🎖 Honorable Mention: Most Impactful – $20,000 in cash (1 winner)
💡 Best Use of Slack Prize – $25,000 in cash (1 winner)
🎨 Honorable Mention: Most Creative – $20,000 in cash (1 winner)
🎤 Honorable Mention: Best Demo Delivery – $20,000 in cash (1 winner)
🗳 People’s Choice (Public Voting) – $5,000 in cash (1 winner)

If you’re up for the challenge, start preparing now! Have questions or insights? Drop them in the comments!

Explore related content:

TDX 2025 News: Salesforce Agentforce 2dx and AgentExchange

How to Get Your AI-Powered Enhanced Developer Org with Agentforce and Data Cloud

New Agentforce Specialist Certification

Your Ultimate Guide to Dreamforce 2024

#Agentforce #Hackathon #Salesforce #SalesforceAdmin #SalesforceDeveloper #TDX25 #TrailblazerDX

How To Build Custom Flow Actions For Agentforce – Planning Phase

Salesforce Agentforce is a framework that enhances customer service and sales by integrating Salesforce’s powerful capabilities with AI agents. Agentforce accelerates workflows, enhances KPIs and saves manual labor.

The user messaging interface you build on the Agentforce platform leverages Artificial Intelligence Natural Language Processing to process customer requests. Agent builders do not need to define every fine branch in the conversation like designing an Interactive Voice Response (IVR) tree; instead, the Atlas reasoning engine handles these variations by evaluating customer requests (utterances) and matching them to topics. Once a topic is matched, Agentforce triggers the relevant actions, executes the predefined instructions, and completes the transaction for the customer.

In short, Agentforce is a superpowered chatbot that can understand customer utterances that are not necessarily well defined, and completes transactions for the user. Voice interface for Agentforce is scheduled to come in the coming releases.

Agentforce Deployment and Distribution

There are two ways you can deploy Agentforce in Salesforce:

1. Internal: Agentforce Assistant provides a conversational interface that enables internal users to efficiently complete transactions.
2. External: Agentforce can be placed on a digital experience (community) site or on your company’s own website to complete transactions for your customers or prospects.

You can also use Agentforce for both internal and external purposes.

Agentforce Topics

Agentforce topics are predefined or custom conversational subjects within Salesforce Agentforce that enable the platform to understand and categorize customer requests (utterances). These topics serve as the foundation for processing conversations by identifying the intent behind a customer’s input and determining the appropriate actions to address it. Topics can be tailored to specific business needs, allowing organizations to design interactions that align with their workflows and objectives. For example, topics might include updating account details, creating support cases, or processing orders. By leveraging AI and natural language processing, Agentforce responsively matches utterances to the most relevant topic, ensuring accurate and efficient transaction handling.

Agentforce comes with a few out of the box topics, and also a few – around 30 at the time of this post – standard actions you can choose from. The advantage of using out of the box topics and actions is that it is faster to implement. On the other hand, out of the box topics and actions are not tailored to your needs, therefore yielding potentially unpredictable results. Now, let’s explore actions. Remember that you can always build custom topics and actions tailored to your needs.

Agentforce Actions

Salesforce Agentforce offers two primary types of actions to enhance its AI agents’ capabilities:

1. Standard Actions: Pre-configured tasks that automate common functions within Salesforce.
2. Custom Actions: Tailored tasks created to meet specific business needs, which can be implemented using:
   • Flow Actions: Utilize Salesforce Flows to automate processes and guide user interactions.
   • Apex Actions: Employ Apex code for complex logic and integrations.
   • Prompt Template Actions: Generate dynamic responses or trigger specific actions based on user input.
   • Service Catalog Item (Beta):  These Service Catalog actions resolve common contact reasons such as update phone number, update email address and open a support case using the conversational interface.

While standard actions can help you to identify a record by name, query or summarize a record, you will need a flow or Apex action to complete a transaction. I won’t go into the benefits of low code in this article, but let me just say that the flow tool is more accessible for most as a solution in this case.

How Do You Build Topics and Actions in Agent Builder?

Salesforce already comes with an Agent built in. I am assuming you will want to add topics and actions to this Agent to handle your unique requirements. You can also build another agent in addition to the one that comes in your Salesforce Org.

Before you jump into building in the agent builder, let me tell you that this process requires several pieces to be delicately stitched together. The building process reminds me very much of the orchestration building process with the obvious difference of agents requiring much more written instructions than orchestrations.

Therefore it is a very good idea to do some planning before building.

There are two main areas driving your design process:

1. Conversation design: You need to think through and consider how the messaging interaction will go down. This will determine how you will structure your topics and actions.
2. Automation design: This item goes hand-in-hand with the first item. Once you envision how the interaction will happen, you define topics to address the utterances, and associate the actions you build for the topic.

🚨 Use case 👇🏼

When the user asks Agentforce for a list of course cohort participants, the participant list will be displayed. There are courses (parent) and cohort (child) objects in the system, and contacts are enrolled in cohorts with a junction object.

I can easily build a single flow to do this. The flow will need the course as an input, and an earliest start date. The flow will fetch the cohort best matching the requirements and display the associated participant contact information to the user.

Conversation Design

1. User: Give me a list of participants for the course starting right after 02/01/2025
2. Agentforce: Here is the list: 1. John Doe 2. Penny White 3. Rick Smith

1. User: Give me a list of participants for the course starting right after 02/01/2025
2. Agentforce: Here is the cohort I found, is this the one? Code: C-0123 Starting Date: 02/07/2025
3. User: Yes
4. Agentforce: Here is the list: 1. John Doe 2. Penny White 3. Rick Smith

The topics and custom actions needed for this design are completely different.

I can and I should further expand on the possibilities here. What happens if there are multiple cohorts starting closely after the earliest start date? What happens if the user says no? You see the picture here, don’t you?

I’ll explore the intricacies of building my Agentforce custom flow actions in a future post. The topic of Automation Design is significant and deserves its own detailed discussion.

Conclusion

Preparation and conversation design are key when building agents for Salesforce Agentforce. The builder does not perfectly lend itself to easy modifications. When you build a custom flow action, and need to modify it later, you can save and activate a new version of your flow, but the custom action does not see these modifications. There is no refresh functionality for the custom action to fetch the latest version with new inputs and outputs.

Planning your Agentforce solution will save you time when iterating and building it. Design and implementation precision will yield accurate and secure outcomes.

Explore related content:

Salesforce is Offering FREE AI Certification

How to Pass the Salesforce AI Specialist Exam

Introducing: Agentforce

Agentforce Hackathon TDX25: My Hackathon Tips

#Admins #AgentBuilder #Agentforce #AI #Apex #Automation #Developers #LowCode #Salesforce

Is Agentforce a champ or a chump?

This week, @Salesforce announced #AgentForce 2.0. 

How big a thing is this? Is it a full pivot? 

Should other #EnSW vendors be worried? How do #agents relate to humans if a "digital labor platform" becomes a thing?

Here is my assessment. https://zurl.co/rEL49

@Salesforce just released #Agentforce 3. What does this mean for #enterprises and the #EnSW market?

Is it ready for the enterprise or are additional iterations required?

Many questions.

Some answers.

https://zurl.co/huzLr

@Salesforce just released #Agentforce 3. What does this mean for #enterprises and the #EnSW market?

Is it ready for the enterprise or are additional iterations required?

Many questions.

Some answers.

https://zurl.co/huzLr

Is Agentforce a champ or a chump?

This week, @Salesforce announced #AgentForce 2.0. 

How big a thing is this? Is it a full pivot? 

Should other #EnSW vendors be worried? How do #agents relate to humans if a "digital labor platform" becomes a thing?

Here is my assessment. https://zurl.co/rEL49

Is Agentforce a champ or a chump?

This week, @Salesforce announced #AgentForce 2.0. 

How big a thing is this? Is it a full pivot? 

Should other #EnSW vendors be worried? How do #agents relate to humans if a "digital labor platform" becomes a thing?

Here is my assessment. https://zurl.co/rEL49