LlamaIndex Newsletter 2025-05-13：Deep Research 智能体工作坊与 LlamaParse 新功能

Hi there, Loyal Llama League! 🦙

Welcome to this week's edition of the LlamaIndex newsletter! We're thrilled to share exciting updates, including a workshop on building Deep Research agents, new features in LlamaParse, and innovative multi-modal retrieval techniques. Dive into these highlights and more to enhance your experience with LlamaIndex!

💰 FinTech event

We're hosting an exclusive workshop in New York city to bring together the top thinkers, builders, and decision-makers shaping the future of finance at Agents and RAG in Finance. You can be part of it! Sign up here.

Explore our free and paid plans today.

🤩 The Highlights:

• Build Your Own Deep Research Agent: Learn how to create a multi-agent system from scratch in LlamaIndex, covering everything from single-agent workflows to advanced multi-agent systems. Check out the notebook or watch a video.
• Multi-Modal Retrieval Techniques: Discover how to use Voyage AI's multi-modal embeddings and MongoDB's multi-modal indexes for effective image and text data handling. Explore the full notebook here.

🗺️ LlamaParse & LlamaParse:

• LlamaParse Enhancements: Experience the latest features in LlamaParse, including new models, auto orientation detection, confidence scores, and customizable error handling. Read more about these improvements.
• LlamaExtract Features: Enhance your AI applications with citation capabilities and improved reasoning for complex data sources. Read more on our blog.

✨ Framework:

• Agent Workflow in TypeScript: Exciting updates on AgentWorkflow now available in TypeScript for building powerful AI systems. Learn more.

✍️ Community:

• AI Agent Development: Explore how our AI agent can accurately extract information from complex documents. Learn more.
• Case Study on LlamaParse: Discover how 11x reduced SDR ramp time using LlamaParse. Read the full case study.

Thank you for being a part of our community! Stay tuned for more updates and insights from LlamaIndex.