#software-architecture โ Public Fediverse posts
Live and recent posts from across the Fediverse tagged #software-architecture, aggregated by home.social.
-
๐ช๐ต๐ ๐๐ ๐๐ด๐ฒ๐ป๐๐ ๐ก๐ฒ๐ฒ๐ฑ ๐๐ผ๐๐ฒ๐ฟ๐ป๐ฎ๐ป๐ฐ๐ฒ โ ๐๐ป๐๐ฒ๐ฟ๐๐ถ๐ฒ๐ ๐๐ถ๐๐ต ๐ ๐ฎ๐ต๐ฏ๐ผ๐๐ฏ๐ฎ ๐๐ต๐ฎ๐ฟ๐ฏ๐ถ & ๐๐ฟ. ๐ฆรถ๐ป๐ธ๐ฒ ๐ ๐ฎ๐ด๐ป๐๐๐๐ฒ๐ป ๐ค In our latest interview, Mahbouba Gharbi and Dr. Sรถnke Magnussen, curators of the #CPSA Advanced-Level module #SWARC4AI, explain why #AIAgents cannot replace human responsibility, why guardrails alone are not enough, and how software architectures can ensure governance, security, and accountability.
Read the full interview ๐ https://t1p.de/6960m
-
๐ช๐ต๐ ๐๐ ๐๐ด๐ฒ๐ป๐๐ ๐ก๐ฒ๐ฒ๐ฑ ๐๐ผ๐๐ฒ๐ฟ๐ป๐ฎ๐ป๐ฐ๐ฒ โ ๐๐ป๐๐ฒ๐ฟ๐๐ถ๐ฒ๐ ๐๐ถ๐๐ต ๐ ๐ฎ๐ต๐ฏ๐ผ๐๐ฏ๐ฎ ๐๐ต๐ฎ๐ฟ๐ฏ๐ถ & ๐๐ฟ. ๐ฆรถ๐ป๐ธ๐ฒ ๐ ๐ฎ๐ด๐ป๐๐๐๐ฒ๐ป ๐ค In our latest interview, Mahbouba Gharbi and Dr. Sรถnke Magnussen, curators of the #CPSA Advanced-Level module #SWARC4AI, explain why #AIAgents cannot replace human responsibility, why guardrails alone are not enough, and how software architectures can ensure governance, security, and accountability.
Read the full interview ๐ https://t1p.de/6960m
-
๐ช๐ต๐ ๐๐ ๐๐ด๐ฒ๐ป๐๐ ๐ก๐ฒ๐ฒ๐ฑ ๐๐ผ๐๐ฒ๐ฟ๐ป๐ฎ๐ป๐ฐ๐ฒ โ ๐๐ป๐๐ฒ๐ฟ๐๐ถ๐ฒ๐ ๐๐ถ๐๐ต ๐ ๐ฎ๐ต๐ฏ๐ผ๐๐ฏ๐ฎ ๐๐ต๐ฎ๐ฟ๐ฏ๐ถ & ๐๐ฟ. ๐ฆรถ๐ป๐ธ๐ฒ ๐ ๐ฎ๐ด๐ป๐๐๐๐ฒ๐ป ๐ค In our latest interview, Mahbouba Gharbi and Dr. Sรถnke Magnussen, curators of the #CPSA Advanced-Level module #SWARC4AI, explain why #AIAgents cannot replace human responsibility, why guardrails alone are not enough, and how software architectures can ensure governance, security, and accountability.
Read the full interview ๐ https://t1p.de/6960m
-
๐ช๐ต๐ ๐๐ ๐๐ด๐ฒ๐ป๐๐ ๐ก๐ฒ๐ฒ๐ฑ ๐๐ผ๐๐ฒ๐ฟ๐ป๐ฎ๐ป๐ฐ๐ฒ โ ๐๐ป๐๐ฒ๐ฟ๐๐ถ๐ฒ๐ ๐๐ถ๐๐ต ๐ ๐ฎ๐ต๐ฏ๐ผ๐๐ฏ๐ฎ ๐๐ต๐ฎ๐ฟ๐ฏ๐ถ & ๐๐ฟ. ๐ฆรถ๐ป๐ธ๐ฒ ๐ ๐ฎ๐ด๐ป๐๐๐๐ฒ๐ป ๐ค In our latest interview, Mahbouba Gharbi and Dr. Sรถnke Magnussen, curators of the #CPSA Advanced-Level module #SWARC4AI, explain why #AIAgents cannot replace human responsibility, why guardrails alone are not enough, and how software architectures can ensure governance, security, and accountability.
Read the full interview ๐ https://t1p.de/6960m
-
We just published *Cutting Your Events the Right Way*, the first part of a two-part series on event design.
It takes the same customer data, models it as one big event and as several small ones, and works out when each cut is the right call.
Read it here: https://patchlevel.dev/blog/cutting-your-events-the-right-way
#EventSourcing #DDD #PHP #SoftwareArchitecture #EventDriven #TechBlog
-
We just published *Cutting Your Events the Right Way*, the first part of a two-part series on event design.
It takes the same customer data, models it as one big event and as several small ones, and works out when each cut is the right call.
Read it here: https://patchlevel.dev/blog/cutting-your-events-the-right-way
#EventSourcing #DDD #PHP #SoftwareArchitecture #EventDriven #TechBlog
-
We just published *Cutting Your Events the Right Way*, the first part of a two-part series on event design.
It takes the same customer data, models it as one big event and as several small ones, and works out when each cut is the right call.
Read it here: https://patchlevel.dev/blog/cutting-your-events-the-right-way
#EventSourcing #DDD #PHP #SoftwareArchitecture #EventDriven #TechBlog
-
We just published *Cutting Your Events the Right Way*, the first part of a two-part series on event design.
It takes the same customer data, models it as one big event and as several small ones, and works out when each cut is the right call.
Read it here: https://patchlevel.dev/blog/cutting-your-events-the-right-way
#EventSourcing #DDD #PHP #SoftwareArchitecture #EventDriven #TechBlog
-
AI agents can recreate the old rockstar developer problem: impressive code that only one mind can explain.
The fix is shared design, local repo rules, small diffs, tests, and review that asks whether the change belongs.
https://www.the-main-thread.com/p/ai-agents-rockstar-developer-problem
-
AI agents can recreate the old rockstar developer problem: impressive code that only one mind can explain.
The fix is shared design, local repo rules, small diffs, tests, and review that asks whether the change belongs.
https://www.the-main-thread.com/p/ai-agents-rockstar-developer-problem
-
AI agents can recreate the old rockstar developer problem: impressive code that only one mind can explain.
The fix is shared design, local repo rules, small diffs, tests, and review that asks whether the change belongs.
https://www.the-main-thread.com/p/ai-agents-rockstar-developer-problem
-
AI agents can recreate the old rockstar developer problem: impressive code that only one mind can explain.
The fix is shared design, local repo rules, small diffs, tests, and review that asks whether the change belongs.
https://www.the-main-thread.com/p/ai-agents-rockstar-developer-problem
-
AI agents can recreate the old rockstar developer problem: impressive code that only one mind can explain.
The fix is shared design, local repo rules, small diffs, tests, and review that asks whether the change belongs.
https://www.the-main-thread.com/p/ai-agents-rockstar-developer-problem
-
Why pure cloud fails in IoT. โ๏ธโก๏ธ Hybrid architecture is the way to go. #tech #iot #cloudcomputing #softwarearchitecture
-
Why pure cloud fails in IoT. โ๏ธโก๏ธ Hybrid architecture is the way to go. #tech #iot #cloudcomputing #softwarearchitecture
-
Why pure cloud fails in IoT. โ๏ธโก๏ธ Hybrid architecture is the way to go. #tech #iot #cloudcomputing #softwarearchitecture
-
Why pure cloud fails in IoT. โ๏ธโก๏ธ Hybrid architecture is the way to go. #tech #iot #cloudcomputing #softwarearchitecture
-
Why pure cloud fails in IoT. โ๏ธโก๏ธ Hybrid architecture is the way to go. #tech #iot #cloudcomputing #softwarearchitecture
-
Been playing with different AI agent orchestration patterns. Event-driven vs scheduled polling โ each has clear tradeoffs.
Event-driven: lower latency, complex state
Polling: simpler, wastes resourcesWhat's your experience? What patterns work best for multi-agent systems?
-
Been playing with different AI agent orchestration patterns. Event-driven vs scheduled polling โ each has clear tradeoffs.
Event-driven: lower latency, complex state
Polling: simpler, wastes resourcesWhat's your experience? What patterns work best for multi-agent systems?
-
Been playing with different AI agent orchestration patterns. Event-driven vs scheduled polling โ each has clear tradeoffs.
Event-driven: lower latency, complex state
Polling: simpler, wastes resourcesWhat's your experience? What patterns work best for multi-agent systems?
-
Agents do not need RAG or vector databases for most real world work. They need structure and semantics.
Agent Knowledge Graphs turn mixed repositories of code, docs, configs, and PDFs into a connected model that agents can reason over. This often replaces entire retrieval pipelines.
https://antaoalmada.dev/posts/Code-Agent-Knowledge-Graphs/
#AIEngineering #KnowledgeGraphs #CodingAgents #AgentWorkflows #SoftwareArchitecture #Graphify
-
Agents do not need RAG or vector databases for most real world work. They need structure and semantics.
Agent Knowledge Graphs turn mixed repositories of code, docs, configs, and PDFs into a connected model that agents can reason over. This often replaces entire retrieval pipelines.
https://antaoalmada.dev/posts/Code-Agent-Knowledge-Graphs/
#AIEngineering #KnowledgeGraphs #CodingAgents #AgentWorkflows #SoftwareArchitecture #Graphify
-
Agents do not need RAG or vector databases for most real world work. They need structure and semantics.
Agent Knowledge Graphs turn mixed repositories of code, docs, configs, and PDFs into a connected model that agents can reason over. This often replaces entire retrieval pipelines.
https://antaoalmada.dev/posts/Code-Agent-Knowledge-Graphs/
#AIEngineering #KnowledgeGraphs #CodingAgents #AgentWorkflows #SoftwareArchitecture #Graphify
-
Agents do not need RAG or vector databases for most real world work. They need structure and semantics.
Agent Knowledge Graphs turn mixed repositories of code, docs, configs, and PDFs into a connected model that agents can reason over. This often replaces entire retrieval pipelines.
https://antaoalmada.dev/posts/Code-Agent-Knowledge-Graphs/
#AIEngineering #KnowledgeGraphs #CodingAgents #AgentWorkflows #SoftwareArchitecture #Graphify
-
Agents do not need RAG or vector databases for most real world work. They need structure and semantics.
Agent Knowledge Graphs turn mixed repositories of code, docs, configs, and PDFs into a connected model that agents can reason over. This often replaces entire retrieval pipelines.
https://antaoalmada.dev/posts/Code-Agent-Knowledge-Graphs/
#AIEngineering #KnowledgeGraphs #CodingAgents #AgentWorkflows #SoftwareArchitecture #Graphify
-
Saw an article starting with:
"If you run anything in the cloud, your teams are already using AI: ChatGPT plugins, Copilot in integrated development environments (IDEs), a LangChain proof of concept that somehow became part of a customer journey, a weekend assistant that was never meant to last but did."
Really? Has it gotten that much out of hand that people don't know anymore where they shipped some AI stuff? Or is this just BS?
-
Saw an article starting with:
"If you run anything in the cloud, your teams are already using AI: ChatGPT plugins, Copilot in integrated development environments (IDEs), a LangChain proof of concept that somehow became part of a customer journey, a weekend assistant that was never meant to last but did."
Really? Has it gotten that much out of hand that people don't know anymore where they shipped some AI stuff? Or is this just BS?
-
Saw an article starting with:
"If you run anything in the cloud, your teams are already using AI: ChatGPT plugins, Copilot in integrated development environments (IDEs), a LangChain proof of concept that somehow became part of a customer journey, a weekend assistant that was never meant to last but did."
Really? Has it gotten that much out of hand that people don't know anymore where they shipped some AI stuff? Or is this just BS?
-
Saw an article starting with:
"If you run anything in the cloud, your teams are already using AI: ChatGPT plugins, Copilot in integrated development environments (IDEs), a LangChain proof of concept that somehow became part of a customer journey, a weekend assistant that was never meant to last but did."
Really? Has it gotten that much out of hand that people don't know anymore where they shipped some AI stuff? Or is this just BS?
-
#AgenticAI architecture is redefining how software is designed - and it's poised to shape the industry for years to come.
This #InfoQ eMag explores the core concepts, patterns, and emerging trends driving agentic AI as it moves into the mainstream.
๐ฅ Download your free copy here: https://bit.ly/4p0tmyv
-
#AgenticAI architecture is redefining how software is designed - and it's poised to shape the industry for years to come.
This #InfoQ eMag explores the core concepts, patterns, and emerging trends driving agentic AI as it moves into the mainstream.
๐ฅ Download your free copy here: https://bit.ly/4p0tmyv
-
#AgenticAI architecture is redefining how software is designed - and it's poised to shape the industry for years to come.
This #InfoQ eMag explores the core concepts, patterns, and emerging trends driving agentic AI as it moves into the mainstream.
๐ฅ Download your free copy here: https://bit.ly/4p0tmyv
-
#AgenticAI architecture is redefining how software is designed - and it's poised to shape the industry for years to come.
This #InfoQ eMag explores the core concepts, patterns, and emerging trends driving agentic AI as it moves into the mainstream.
๐ฅ Download your free copy here: https://bit.ly/4p0tmyv
-
#AgenticAI architecture is redefining how software is designed - and it's poised to shape the industry for years to come.
This #InfoQ eMag explores the core concepts, patterns, and emerging trends driving agentic AI as it moves into the mainstream.
๐ฅ Download your free copy here: https://bit.ly/4p0tmyv
-
โThe real problem is not whether machines think but whether men do.โ โ B. F. Skinner
New tools, AI included, can accelerate development, but they do not replace critical thinking. Architecture decisions, risk assessment, and system design still require human judgment. Better tools make thinking more important, not less.
๐ฌ How do you ensure that convenience does not replace understanding?
#DevThinking #SoftwareArchitecture #CodeQuality
Photo by Eugen Str on Unsplash
-
โThe real problem is not whether machines think but whether men do.โ โ B. F. Skinner
New tools, AI included, can accelerate development, but they do not replace critical thinking. Architecture decisions, risk assessment, and system design still require human judgment. Better tools make thinking more important, not less.
๐ฌ How do you ensure that convenience does not replace understanding?
#DevThinking #SoftwareArchitecture #CodeQuality
Photo by Eugen Str on Unsplash
-
Target built a new GenAI system to overhaul marketing campaign forecasting by retrieving and ranking similar historical campaigns.
Instead of relying on rule-based workflows, it uses embeddings, vector search, and LLM-based ranking to identify the most relevant past campaigns.
Read #InfoQ to see how the system performed ๐ https://bit.ly/4p0w8Uh
-
Target built a new GenAI system to overhaul marketing campaign forecasting by retrieving and ranking similar historical campaigns.
Instead of relying on rule-based workflows, it uses embeddings, vector search, and LLM-based ranking to identify the most relevant past campaigns.
Read #InfoQ to see how the system performed ๐ https://bit.ly/4p0w8Uh
-
Target built a new GenAI system to overhaul marketing campaign forecasting by retrieving and ranking similar historical campaigns.
Instead of relying on rule-based workflows, it uses embeddings, vector search, and LLM-based ranking to identify the most relevant past campaigns.
Read #InfoQ to see how the system performed ๐ https://bit.ly/4p0w8Uh
-
Target built a new GenAI system to overhaul marketing campaign forecasting by retrieving and ranking similar historical campaigns.
Instead of relying on rule-based workflows, it uses embeddings, vector search, and LLM-based ranking to identify the most relevant past campaigns.
Read #InfoQ to see how the system performed ๐ https://bit.ly/4p0w8Uh
-
Target built a new GenAI system to overhaul marketing campaign forecasting by retrieving and ranking similar historical campaigns.
Instead of relying on rule-based workflows, it uses embeddings, vector search, and LLM-based ranking to identify the most relevant past campaigns.
Read #InfoQ to see how the system performed ๐ https://bit.ly/4p0w8Uh
-
Microservices keep getting recommended to teams that don't yet have a working monolith... Backwards!
Most successful microservice stories started as monoliths that got broken up. Architecture depends on the problem, not the conference talk.
-
Microservices keep getting recommended to teams that don't yet have a working monolith... Backwards!
Most successful microservice stories started as monoliths that got broken up. Architecture depends on the problem, not the conference talk.
-
Event-driven architecture promises scalability, but the real tradeoffs in Java-based real-time systems only show up in production.
Drawing on a Java/Kafka contact center platform handling 80k BHCC across 10k agents, Sagar Deepak Joshi's new #InfoQ article explores exactly where things break down:
๐น State management & partition limits
๐น Message deduplication
๐น JVM tuning challenges
๐น Cascading consumer failuresDiscover the Redis-backed patterns used to solve them and keep the system resilient.
๐ Read now for more insights: https://bit.ly/4bmaRPb
#Java #SpringBoot #ApacheKafka #Redis #Microservices #SoftwareArchitecture
-
Event-driven architecture promises scalability, but the real tradeoffs in Java-based real-time systems only show up in production.
Drawing on a Java/Kafka contact center platform handling 80k BHCC across 10k agents, Sagar Deepak Joshi's new #InfoQ article explores exactly where things break down:
๐น State management & partition limits
๐น Message deduplication
๐น JVM tuning challenges
๐น Cascading consumer failuresDiscover the Redis-backed patterns used to solve them and keep the system resilient.
๐ Read now for more insights: https://bit.ly/4bmaRPb
#Java #SpringBoot #ApacheKafka #Redis #Microservices #SoftwareArchitecture
-
Event-driven architecture promises scalability, but the real tradeoffs in Java-based real-time systems only show up in production.
Drawing on a Java/Kafka contact center platform handling 80k BHCC across 10k agents, Sagar Deepak Joshi's new #InfoQ article explores exactly where things break down:
๐น State management & partition limits
๐น Message deduplication
๐น JVM tuning challenges
๐น Cascading consumer failuresDiscover the Redis-backed patterns used to solve them and keep the system resilient.
๐ Read now for more insights: https://bit.ly/4bmaRPb
#Java #SpringBoot #ApacheKafka #Redis #Microservices #SoftwareArchitecture
-
Event-driven architecture promises scalability, but the real tradeoffs in Java-based real-time systems only show up in production.
Drawing on a Java/Kafka contact center platform handling 80k BHCC across 10k agents, Sagar Deepak Joshi's new #InfoQ article explores exactly where things break down:
๐น State management & partition limits
๐น Message deduplication
๐น JVM tuning challenges
๐น Cascading consumer failuresDiscover the Redis-backed patterns used to solve them and keep the system resilient.
๐ Read now for more insights: https://bit.ly/4bmaRPb
#Java #SpringBoot #ApacheKafka #Redis #Microservices #SoftwareArchitecture
-
Event-driven architecture promises scalability, but the real tradeoffs in Java-based real-time systems only show up in production.
Drawing on a Java/Kafka contact center platform handling 80k BHCC across 10k agents, Sagar Deepak Joshi's new #InfoQ article explores exactly where things break down:
๐น State management & partition limits
๐น Message deduplication
๐น JVM tuning challenges
๐น Cascading consumer failuresDiscover the Redis-backed patterns used to solve them and keep the system resilient.
๐ Read now for more insights: https://bit.ly/4bmaRPb
#Java #SpringBoot #ApacheKafka #Redis #Microservices #SoftwareArchitecture
-
Most issues with coding harnesses arenโt about capabilityโtheyโre about structure.
A useful framing:
โข Instructions = intent (what should be done)
โข Skills = capabilities (what can be done)
โข Agents = orchestration (how itโs done)
โข Hooks = control points (where you intervene)Keep these concerns separate, and harnesses become easier to reason about, extend, and debug. Mix them, and complexity compounds fast.
https://antaoalmada.dev/posts/The-Architecture-of-Code-Agents-Instructions-Skills-Agents-Hooks/
-
Most issues with coding harnesses arenโt about capabilityโtheyโre about structure.
A useful framing:
โข Instructions = intent (what should be done)
โข Skills = capabilities (what can be done)
โข Agents = orchestration (how itโs done)
โข Hooks = control points (where you intervene)Keep these concerns separate, and harnesses become easier to reason about, extend, and debug. Mix them, and complexity compounds fast.
https://antaoalmada.dev/posts/The-Architecture-of-Code-Agents-Instructions-Skills-Agents-Hooks/
-
Most issues with coding harnesses arenโt about capabilityโtheyโre about structure.
A useful framing:
โข Instructions = intent (what should be done)
โข Skills = capabilities (what can be done)
โข Agents = orchestration (how itโs done)
โข Hooks = control points (where you intervene)Keep these concerns separate, and harnesses become easier to reason about, extend, and debug. Mix them, and complexity compounds fast.
https://antaoalmada.dev/posts/The-Architecture-of-Code-Agents-Instructions-Skills-Agents-Hooks/
-
Most issues with coding harnesses arenโt about capabilityโtheyโre about structure.
A useful framing:
โข Instructions = intent (what should be done)
โข Skills = capabilities (what can be done)
โข Agents = orchestration (how itโs done)
โข Hooks = control points (where you intervene)Keep these concerns separate, and harnesses become easier to reason about, extend, and debug. Mix them, and complexity compounds fast.
https://antaoalmada.dev/posts/The-Architecture-of-Code-Agents-Instructions-Skills-Agents-Hooks/
-
Most issues with coding harnesses arenโt about capabilityโtheyโre about structure.
A useful framing:
โข Instructions = intent (what should be done)
โข Skills = capabilities (what can be done)
โข Agents = orchestration (how itโs done)
โข Hooks = control points (where you intervene)Keep these concerns separate, and harnesses become easier to reason about, extend, and debug. Mix them, and complexity compounds fast.
https://antaoalmada.dev/posts/The-Architecture-of-Code-Agents-Instructions-Skills-Agents-Hooks/
-
Not every pattern improves a system. A design pattern applied without understanding often adds more complexity than value. The best pattern is the one that solves a real problem and disappears into the background.
Based on โPatterns Compactโ by Karl Eilebrecht. #SoftwareArchitecture #DesignPatterns #Engineering
Photo by Amsterdam City Archives on Unsplash
-
Not every pattern improves a system. A design pattern applied without understanding often adds more complexity than value. The best pattern is the one that solves a real problem and disappears into the background.
Based on โPatterns Compactโ by Karl Eilebrecht. #SoftwareArchitecture #DesignPatterns #Engineering
Photo by Amsterdam City Archives on Unsplash
-
Imagine inventing something, watching it become an industry standard, and then spending decades apologizing for it.
That's exactly what computing pioneer Sir Tony Hoare did in his legendary talk, "Null References: The Billion-Dollar Mistakeโ.
As part of the #InfoQ20 campaign, we're revisiting this classic archive video - a must-watch for every software practitioner. It's a powerful lesson in how a seemingly small design decision can shape the entire software industry for decades.
๐ฌ Watch the classic talk here: https://bit.ly/4aotulm
#ComputerScience #SoftwareEngineering #SoftwareArchitecture #TechHistory
-
Imagine inventing something, watching it become an industry standard, and then spending decades apologizing for it.
That's exactly what computing pioneer Sir Tony Hoare did in his legendary talk, "Null References: The Billion-Dollar Mistakeโ.
As part of the #InfoQ20 campaign, we're revisiting this classic archive video - a must-watch for every software practitioner. It's a powerful lesson in how a seemingly small design decision can shape the entire software industry for decades.
๐ฌ Watch the classic talk here: https://bit.ly/4aotulm
#ComputerScience #SoftwareEngineering #SoftwareArchitecture #TechHistory
-
Imagine inventing something, watching it become an industry standard, and then spending decades apologizing for it.
That's exactly what computing pioneer Sir Tony Hoare did in his legendary talk, "Null References: The Billion-Dollar Mistakeโ.
As part of the #InfoQ20 campaign, we're revisiting this classic archive video - a must-watch for every software practitioner. It's a powerful lesson in how a seemingly small design decision can shape the entire software industry for decades.
๐ฌ Watch the classic talk here: https://bit.ly/4aotulm
#ComputerScience #SoftwareEngineering #SoftwareArchitecture #TechHistory
-
Imagine inventing something, watching it become an industry standard, and then spending decades apologizing for it.
That's exactly what computing pioneer Sir Tony Hoare did in his legendary talk, "Null References: The Billion-Dollar Mistakeโ.
As part of the #InfoQ20 campaign, we're revisiting this classic archive video - a must-watch for every software practitioner. It's a powerful lesson in how a seemingly small design decision can shape the entire software industry for decades.
๐ฌ Watch the classic talk here: https://bit.ly/4aotulm
#ComputerScience #SoftwareEngineering #SoftwareArchitecture #TechHistory