#llamaindex — Public Fediverse posts

RAG для закупок: Qdrant и LlamaIndex в локальном контуре

Привет Хабр! Меня зовут Владимир, и недавно я решил изучить новую (для себя) технологию - LlamaIndex . А тут и задачка подвернулась - надоело копаться в Положении о закупках, поэтому понадобился RAG для ответов по ФЗ-44, ФЗ-223, ну и локальному положению. В этой статье разберу, как создать простенький RAG, не выходящий из локального контура, на базе LlamaIndex + Qdrant, напишем к нему API и UI на Gradio. Поехали.

https://habr.com/ru/articles/1056194/

#qdrant #llamaindex #rag #python #векторный_поиск #llm #fastapi

Агентные фреймворки: обещали революцию, что осталось в 2026

Два года назад все хотели агента, который «сам пишет код, сам его тестирует и сам деплоит в прод». Сейчас 2026 год. Давайте честно поговорим о том, что из этого взлетело, а что тихо умерло в корпоративных slack-каналах под сообщениями «окей, пока притормозим с агентами».

https://habr.com/ru/articles/1041274/

#LLMагенты #LangChain #LangGraph #AutoGen #MCP #RAG #агентные_системы #AI_в_продакшене #LlamaIndex

We shipped a GenAI card generator for the EO Philadelphia chapter in days, not months. FastAPI + LlamaIndex + GPT-4o-mini + Gemini 3 Pro, turning member photos into branded New Year cards. Proof that small-scope AI tools can land real community impact. https://go.ombulabs.com/jh9 #GenAI #LlamaIndex #FastAPI #Entrepreneurship

RE: https://fedi.simonwillison.net/@simon/116457708120212477

#liteparse werde ich im Vergleich zu #Docling lokal mal testen, nutze das bisher nur per Web über #llamaindex.

RE: https://fedi.simonwillison.net/@simon/116457708120212477

#liteparse werde ich im Vergleich zu #Docling lokal mal testen, nutze das bisher nur per Web über #llamaindex.

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

como con APIs cloud; el perfil de privacidad depende del backend elegido.

Facilidad de uso
7/10
Capacidad analítica
7/10
Multimodalidad
5/10
Privacidad
9/10
Integraciones
7/10
Ejecución local
9/10
Sitio oficial https://anythingllm.com/

3) #LlamaIndex
Presenta una fortaleza alta en investigación y búsqueda.

Para construir RAG a medida con control total sobre ingesta, #indexación, recuperación y generación; requiere capacidad técnica

https://www.tkhunt.com/2291331/ RAGの主流となった「Agentic RAG（エージェント型RAG）」、徹底解説 #AgenticAi #AgenticRAG #AI #AIエージェント #ArtificialIntelligence #Cohere #ContextualAI #LangChain #LlamaIndex #llm #MCP #Pinecone #rag #エージェント型AI #エージェント型RAG #エンタープライズAI #シリコンバレーVC #ベクトルデータベース #人工知能 #生成AI

https://www.tkhunt.com/2291331/ RAGの主流となった「Agentic RAG（エージェント型RAG）」、徹底解説 #AgenticAi #AgenticRAG #AI #AIエージェント #ArtificialIntelligence #Cohere #ContextualAI #LangChain #LlamaIndex #llm #MCP #Pinecone #rag #エージェント型AI #エージェント型RAG #エンタープライズAI #シリコンバレーVC #ベクトルデータベース #人工知能 #生成AI

Hybrid RAG для бизнеса: умный поиск по документам без облака и утечки данных

Крупная логистическая компания. Многолетний архив — договоры с перевозчиками, регламенты, переписка по инцидентам. Менеджер хочет узнать: были ли у нас претензии к этому контрагенту и на каких условиях мы с ним работали два года назад? Открывает почту, потом SharePoint, потом звонит в архив. Через час у него есть частичный ответ. Та же ситуация повторяется в таможне, банках и юридических фирмах. Данные накоплены — в PDF, Word, сканах, таблицах. Но корпоративная база знаний не отвечает на вопросы: нет инструмента, который понимал бы смысл запроса, а не просто искал по ключевым словам. Автоматизация документооборота упёрлась в потолок классических СЭД.

https://habr.com/ru/articles/1024696/

#Автоматизация_Документооборота #ИИ_для_Бизнеса #Нейросеть_Без_Облака #HybridRAG #Self_Hosted #LangGraph #Haystack #Qdrant #LlamaIndex #observability

🚀 44 skills to make any AI coding assistant powerful:

🎮 Games: #pygame, #OpenRCT2, mGBA

🖥️ Desktop: #PyQt, #Qt C++

🌐 Extensions: #Firefox, #Thunderbird

🐍 Python: #Django, pytest, SQLAlchemy

🐧 Linux: #KDE Plasmoid, Kate

🤖 AI: #LlamaIndex

and more!

https://github.com/codeatcode/oss-ai-skills

🚀 44 skills to make any AI coding assistant powerful:

🎮 Games: #pygame, #OpenRCT2, mGBA

🖥️ Desktop: #PyQt, #Qt C++

🌐 Extensions: #Firefox, #Thunderbird

🐍 Python: #Django, pytest, SQLAlchemy

🐧 Linux: #KDE Plasmoid, Kate

🤖 AI: #LlamaIndex

and more!

https://github.com/codeatcode/oss-ai-skills

Анализ документов нейросетью с цитатами из источников: research-docs скилл для Claude Code

Закинул 30 PDF в папку — получил HTML-отчёт с ответами на вопросы и цитатами, привязанными к конкретному месту на странице документа. Research-docs — Claude Code скилл от LlamaIndex на базе парсера LiteParse. Разбираю как работает, какие форматы поддерживает, сколько стоит и где подвох с «никаких галлюцинаций».

https://habr.com/ru/articles/1021098/

#claudecode #pdf #llm #rag #llamaindex #анализ_документов #opensource

Agentic RAG Challenge. Я знаю что вы искали прошлым летом…

В данной статье хотел бы поделиться опытом участия в хакатоне Agentic Legal RAG Challenge 2026 . Наша команда называется "Sparks of intelligence".

https://habr.com/ru/articles/1014520/

#rag #qdrant #llamaindex #ии_агент #векторный_поиск #чанкинг #гибридный_поиск #векторная_база_данных #хакатон

LlamaIndex launches LiteParse, a new TypeScript-native library for parsing complex PDFs in AI agent workflows. The open-source tool runs locally, preserving document layout for better LLM understanding. It addresses a key bottleneck in RAG systems where data ingestion has become slower than the models themselves. https://www.marktechpost.com/2026/03/19/llamaindex-releases-liteparse-a-cli-and-typescript-native-library-for-spatial-pdf-parsing-in-ai-agent-workflows/ #AIagent #AI #GenAI #AgenticAI #LlamaIndex

LlamaIndex launches LiteParse, a new TypeScript-native library for parsing complex PDFs in AI agent workflows. The open-source tool runs locally, preserving document layout for better LLM understanding. It addresses a key bottleneck in RAG systems where data ingestion has become slower than the models themselves. https://www.marktechpost.com/2026/03/19/llamaindex-releases-liteparse-a-cli-and-typescript-native-library-for-spatial-pdf-parsing-in-ai-agent-workflows/ #AIagent #AI #GenAI #AgenticAI #LlamaIndex

LlamaIndex has released LiteParse, a new open-source TypeScript library for parsing PDFs in AI agent workflows. Unlike traditional tools that convert to Markdown, LiteParse preserves spatial layout using a grid approach. Zero Python dependencies, runs locally on Node.js. Aims to solve the data ingestion bottleneck in RAG pipelines. #AIagent #AI #GenAI #AgenticAI #LlamaIndex https://www.marktechpost.com/2026/03/19/llamaindex-releases-liteparse-a-cli-and-typescript-native-library-for-spatial-pdf-parsing-in-ai-agent-workflows/

LlamaIndex has released LiteParse, a new open-source TypeScript library for parsing PDFs in AI agent workflows. Unlike traditional tools that convert to Markdown, LiteParse preserves spatial layout using a grid approach. Zero Python dependencies, runs locally on Node.js. Aims to solve the data ingestion bottleneck in RAG pipelines. #AIagent #AI #GenAI #AgenticAI #LlamaIndex https://www.marktechpost.com/2026/03/19/llamaindex-releases-liteparse-a-cli-and-typescript-native-library-for-spatial-pdf-parsing-in-ai-agent-workflows/

New AI Leader Series 🎥

@Experian_US 's Head of AI/ML Innovation James Lin shares why his team chose LlamaIndex to build AI customer support agents designed to improve NPS — and what mattered most in production.

Watch → https://www.youtube.com/watch?v=bz50n9jDP2M

#AIAgents #EnterpriseAI #LlamaIndex

New AI Leader Series 🎥

@Experian_US 's Head of AI/ML Innovation James Lin shares why his team chose LlamaIndex to build AI customer support agents designed to improve NPS — and what mattered most in production.

Watch → https://www.youtube.com/watch?v=bz50n9jDP2M

#AIAgents #EnterpriseAI #LlamaIndex

My demo project to RAG your email with @regolo_ai is here :-) #ai #llamaindex #python
https://regolo.ai/privacy-first-email-search-building-a-rag-system-with-llamaindex/

My demo project to RAG your email with @regolo_ai is here :-) #ai #llamaindex #python
https://regolo.ai/privacy-first-email-search-building-a-rag-system-with-llamaindex/

Maxun v0.0.32 ra mắt với tính năng ghi âm thời gian thực, hỗ trợ đồng bộ trạng thái website thực tế, thao tác live như gõ, nhấn, cuộn, điều hướng. Hỗ trợ tích hợp SDK: LlamaIndex, Google Sheets, Airtable, LangChain, OpenAI và nhiều hơn nữa. Chế độ AI tự động tìm và trích xuất dữ liệu mà không cần URL. Mã nguồn mở, tự lưu trữ. #Maxun #WebScraping #OpenSource #SelfHosted #AI #LlamaIndex #LangChain #NoCode #DataExtraction #CôngCụMãNguồnMở #TríchXuấtDữLiệu #AI #TựHost

https://www.reddit.com/r/selfh

Maxun v0.0.32 ra mắt với tính năng ghi âm thời gian thực, hỗ trợ đồng bộ trạng thái website thực tế, thao tác live như gõ, nhấn, cuộn, điều hướng. Hỗ trợ tích hợp SDK: LlamaIndex, Google Sheets, Airtable, LangChain, OpenAI và nhiều hơn nữa. Chế độ AI tự động tìm và trích xuất dữ liệu mà không cần URL. Mã nguồn mở, tự lưu trữ. #Maxun #WebScraping #OpenSource #SelfHosted #AI #LlamaIndex #LangChain #NoCode #DataExtraction #CôngCụMãNguồnMở #TríchXuấtDữLiệu #AI #TựHost

https://www.reddit.com/r/selfh

LlamaIndex is sponsoring + speaking at @wandb ( @CoreWeave ) Fully Connected London (Nov 4–5), a 2-day event for engineers shipping AI to prod.

Talk: Automating knowledge work with AI agents
Speaker: @tuana (Sr. DevRel Eng)

https://wandb.ai/site/resources/events/fully-connected/london/

#AI #LLM #Agents #LlamaIndex #FullyConnected2025

LlamaIndex is sponsoring + speaking at @wandb ( @CoreWeave ) Fully Connected London (Nov 4–5), a 2-day event for engineers shipping AI to prod.

Talk: Automating knowledge work with AI agents
Speaker: @tuana (Sr. DevRel Eng)

https://wandb.ai/site/resources/events/fully-connected/london/

#AI #LLM #Agents #LlamaIndex #FullyConnected2025

We’re bringing AI agents to the insurance frontier at #ITCVegas2025 🦙
See how LlamaIndex is helping insurers streamline claims, underwriting & CX.

📅 Pre-book a meeting → get exclusive LlamaIndex swag 👇
🔗 https://landing.llamaindex.ai/itcvegas2025

#AIinInsurance #InsurTech #LlamaIndex

We’re bringing AI agents to the insurance frontier at #ITCVegas2025 🦙
See how LlamaIndex is helping insurers streamline claims, underwriting & CX.

📅 Pre-book a meeting → get exclusive LlamaIndex swag 👇
🔗 https://landing.llamaindex.ai/itcvegas2025

#AIinInsurance #InsurTech #LlamaIndex

[Перевод] Автоматический парсинг чеков с LlamaIndex и Pydantic

Команда Python for Devs подготовила перевод статьи о том, как с помощью LlamaIndex и Pydantic можно превратить сканы чеков в структурированные данные. Минимум кода — и у вас готовый CSV для анализа.

https://habr.com/ru/articles/953414/

#Python #LlamaIndex #Pydantic #OCR #Kaggle #Receipt #Extraction #CSV #Async #Pipeline

Discover how Flowise helps you create and build custom AI agents and LLM workflows with a simple drag-and-drop interface.

More details here: https://ostechnix.com/flowise-build-ai-agents-llm-workflows/

#Flowise #AI #LLM #RAG #Langchain #Langgraph #Llamaindex #Opensource #Nodejs

Discover how Flowise helps you create and build custom AI agents and LLM workflows with a simple drag-and-drop interface.

More details here: https://ostechnix.com/flowise-build-ai-agents-llm-workflows/

#Flowise #AI #LLM #RAG #Langchain #Langgraph #Llamaindex #Opensource #Nodejs

📢 Episode 2 of the AI Leader Series is live!

We talk with Swami Chandrasekaran, Head of AI & Data Labs at @KPMG_US, about how the Big Four firm powers context-aware AI agents with LlamaIndex.

👉 Watch now + subscribe: https://www.youtube.com/watch?v=qkpUo9RheOo

#AILeaderSeries #KPMG #EnterpriseAI #LlamaIndex

📢 Episode 2 of the AI Leader Series is live!

We talk with Swami Chandrasekaran, Head of AI & Data Labs at @KPMG_US, about how the Big Four firm powers context-aware AI agents with LlamaIndex.

👉 Watch now + subscribe: https://www.youtube.com/watch?v=qkpUo9RheOo

#AILeaderSeries #KPMG #EnterpriseAI #LlamaIndex

How @Jeppesen (a @Boeing company) went from 512h → 64h to build AI agents:

✅ Built a Unified Chatbot Framework on LlamaIndex
✅ 1,792h saved already
✅ Nearly 4,900h projected annually

From chatbot to full agent orchestration system. 🚀
Case study: https://www.llamaindex.ai/blog/jeppesen-a-boeing-company-saves-2-000-engineering-hours-with-unified-chat-framework-built-on

#GenAI #AIagents #LlamaIndex

How @Jeppesen (a @Boeing company) went from 512h → 64h to build AI agents:

✅ Built a Unified Chatbot Framework on LlamaIndex
✅ 1,792h saved already
✅ Nearly 4,900h projected annually

From chatbot to full agent orchestration system. 🚀
Case study: https://www.llamaindex.ai/blog/jeppesen-a-boeing-company-saves-2-000-engineering-hours-with-unified-chat-framework-built-on

#GenAI #AIagents #LlamaIndex

🚀 Excited to co-sponsor Agentic AI in Action in SF with AWS, Elastic, and TwelveLabs on Aug 26 in SF!

Catch our own Laurie Voss (VP Dev Rel) with a live tech talk: “Building Document Agents with LlamaIndex: Effective Design Patterns”.

Expect food, demos, partner insights, and AI networking magic. See you at the AWS Loft!

🔗 Register today: https://www.aicamp.ai/event/eventdetails/W2025082609

#AgenticAI #LlamaIndex #AWS #Elastic #AI

🚀 Excited to co-sponsor Agentic AI in Action in SF with AWS, Elastic, and TwelveLabs on Aug 26 in SF!

Catch our own Laurie Voss (VP Dev Rel) with a live tech talk: “Building Document Agents with LlamaIndex: Effective Design Patterns”.

Expect food, demos, partner insights, and AI networking magic. See you at the AWS Loft!

🔗 Register today: https://www.aicamp.ai/event/eventdetails/W2025082609

#AgenticAI #LlamaIndex #AWS #Elastic #AI

🚀 @SkySQL just cracked the code on hallucination-free SQL generation.

Using @llamaindex, they built AI agents that turn natural language into accurate SQL queries across complex database schemas.

Key wins:

✅ Zero hallucinated queries
✅ Faster development cycles
✅ Seamless MariaDB integration

The future of database interactions is here 👇
https://www.llamaindex.ai/blog/how-skysql-enables-smarter-text-to-sql-agents-with-llamaindex?utm_source=socials&utm_medium=li_social

#AI #Database #LlamaIndex

🚀 @SkySQL just cracked the code on hallucination-free SQL generation.

Using @llamaindex, they built AI agents that turn natural language into accurate SQL queries across complex database schemas.

Key wins:

✅ Zero hallucinated queries
✅ Faster development cycles
✅ Seamless MariaDB integration

The future of database interactions is here 👇
https://www.llamaindex.ai/blog/how-skysql-enables-smarter-text-to-sql-agents-with-llamaindex?utm_source=socials&utm_medium=li_social

#AI #Database #LlamaIndex

Works with #Python #JavaScript #TypeScript #Go supporting frameworks like #LangChain #LlamaIndex #Genkit & more
Github: https://github.com/googleapis/genai-toolbox
MCP Manual: https://cloud.google.com/sql/docs/mysql/pre-built-tools-with-mcp-toolbox

Works with #Python #JavaScript #TypeScript #Go supporting frameworks like #LangChain #LlamaIndex #Genkit & more
Github: https://github.com/googleapis/genai-toolbox
MCP Manual: https://cloud.google.com/sql/docs/mysql/pre-built-tools-with-mcp-toolbox

How do you trust AI when it extracts data from documents?

That’s been one of the biggest blockers to using AI in high-stakes business settings. If you can’t trace where the information came from… it’s hard to rely on it.

💥 Enter LlamaExtract — a new tool from LlamaIndex that extracts data with citations.

Now when AI pulls out key insights from a document, you can see exactly where it came from.

Read more here: https://www.llamaindex.ai/blog/get-citations-and-reasoning-for-extracted-data-in-llamaextract

#AI #GenAI #GenerativeAI #LlamaIndex

How do you trust AI when it extracts data from documents?

That’s been one of the biggest blockers to using AI in high-stakes business settings. If you can’t trace where the information came from… it’s hard to rely on it.

💥 Enter LlamaExtract — a new tool from LlamaIndex that extracts data with citations.

Now when AI pulls out key insights from a document, you can see exactly where it came from.

Read more here: https://www.llamaindex.ai/blog/get-citations-and-reasoning-for-extracted-data-in-llamaextract

#AI #GenAI #GenerativeAI #LlamaIndex

Neo4j treibt mit GraphRAG, Vektor-Indizes & Agentic RAG die #KI-Entwicklung voran. Ob #LangChain, #LlamaIndex, #SpringAI oder #VertexAI – das neue Python-Paket und das Model Context Protocol (MCP) verknüpfen Graphdaten nahtlos mit #LLM-Anwendungen.
#Neo4j #GenAI #RAG #GraphQL #Cypher #VectorSearch #AgenticAI
https://www.bigdata-insider.de/leistungssprung-bei-graph-datenbanken-mit-ki-integration-cloud-skalierung-und-terabyte-graphen-a-2307ed20cfaf562a1a0094b712b5be95/

Neo4j treibt mit GraphRAG, Vektor-Indizes & Agentic RAG die #KI-Entwicklung voran. Ob #LangChain, #LlamaIndex, #SpringAI oder #VertexAI – das neue Python-Paket und das Model Context Protocol (MCP) verknüpfen Graphdaten nahtlos mit #LLM-Anwendungen.
#Neo4j #GenAI #RAG #GraphQL #Cypher #VectorSearch #AgenticAI
https://www.bigdata-insider.de/leistungssprung-bei-graph-datenbanken-mit-ki-integration-cloud-skalierung-und-terabyte-graphen-a-2307ed20cfaf562a1a0094b712b5be95/

🚀 Dive into the world of multi-agent orchestration with LlamaIndex's AgentWorkflow! 🧠📈 Learn how it simplifies complex scenarios through an easy-to-understand example. Check out my latest article!
#DataScience #AI #MachineLearning #DeepLearning #LlamaIndex
https://www.dataleadsfuture.com/diving-into-llamaindex-agentworkflow-a-nearly-perfect-multi-agent-orchestration-solution/

Ich will mal ein paar #GenAI #RAG Tests machen und habe überlegt, mit welchem mir bekannten Korpus ich das machen könnte. Nachdem ab Dienstag erstmal dystopische Zeiten bevorstehen, habe ich mich für "Jonas, Der letzte Detektiv" entschieden (https://de.wikipedia.org/wiki/Der_letzte_Detektiv) entschieden, das habe ich in meiner Jugend immer Donnerstag Abend in Bayern 2 gehört 😎

Prozess: Download > Transkript mit #Whisper -> Vorbereitung mit #Llamaindex

GraphRAGでポケモン図鑑の構築に挑戦【LlamaIndex × Neptune 】
https://qiita.com/TKG_KM/items/42c31a852470ae4d28db?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #AWS #bedrock #Neptune #Streamlit #LlamaIndex

🔥 Docling: AI-powered document processing! 🔥
🤯 PDFs & DOCXs in your AI workflow? Docling makes it easy! Converts to markdown & JSON for RAG and more. Blazing fast! 🚀

https://youtu.be/zSCxbqgqeJ8?si=mede5eJL_iGRFwAJ

#AI #Docling #DocumentProcessing #OpenSource #RAG #LlamaIndex #Efficiency #Workflows