Hi there, Llama Enthusiasts! 🦙
Welcome to this week's edition of the LlamaIndex newsletter! We're excited to bring you updates on human-in-the-loop workflows, advanced agentic RAG applications, production-ready deployment solutions, and powerful document processing capabilities. Plus, we've got exciting community events and partnerships to share with you.
🤩 The Highlights:
- Human-in-the-Loop Workflows: Build long-running agents that can pause, wait for user input, and resume seamlessly with LlamaIndex workflows. Perfect for complex, multi-step processes where humans need to stay in control of key decisions. Full tutorial.
- Research Agent with Google Gemini 2.5 Pro: Explore how to build a multi-agent research assistant that can search the web, take notes, write reports, and review research using LlamaIndex workflows and Gemini's tool-using capabilities. Start building.
- Advanced Agentic RAG with AWS: Comprehensive guide showing how to create powerful research tools using LlamaIndex, Mistral Large 2, and AWS that integrate with external APIs and internal knowledge bases. Implementation guide.
🗺️ LlamaCloud And LlamaParse:
- One-Click Table Extraction: LlamaParse now supports seamless table extraction from PDFs including financial reports, research papers, and invoices. Check out the notebook.
- NotebookLlaMa Updates: Our open-source NotebookLM alternative now features image and table extraction, interactive data visualization, and claim verification with citation finding. Star on GitHub.
✨ Framework:
- Structured Output Support: All agents and workflows now support structured outputs using Pydantic models for consistent response formats and reliable multi-agent communication. Full tutorial.
- UiPath Enterprise Integration: Deploy LlamaIndex agents seamlessly into enterprise environments with full code-level control, enterprise-grade governance, and single CLI command deployment. Read announcement | Explore SDK.
- Production RAG Guide: Comprehensive hitchhiker's guide covering text extraction strategies, smart chunking approaches, hybrid search techniques, and performance optimization for production-ready RAG systems. Read the guide.
✍️ Community:
- Amsterdam Data Agent Builders Meetup: Partnership with Snowflake featuring hands-on talks about building production-grade data agents with real enterprise data (July 31st). Register.
- LeSearch Research Tool: Multi-agent research assistant built with ReActAgent framework that won Best Use of LlamaIndex at Carnegie Mellon AI Agents Hackathon. Features multi-hop QA, code navigation, and environment dependency resolution. Try LeSearch | Watch demo.
