Hi there, Llama Fans! 🦙
Welcome to this week's edition of the LlamaIndex newsletter! We're thrilled to share some exciting updates, including new tools for building agentic solutions, a powerful multilingual RAG system, and innovative applications for document processing. Dive into the highlights and community contributions that showcase the versatility of LlamaIndex!
🤩 The Highlights:
- PapersChat: An agentic AI application that lets you chat with your papers and gather information from Arxiv and PubMed. It indexes all your papers and provides a user-friendly web UI for querying. Plus, it's open source! Check it out here.
- Multilingual, Multimodal RAG System: Learn how to create a powerful Retrieval-Augmented Generation system that handles multiple languages and modalities seamlessly. Learn more.
- Deep Researcher Template: Use create-llama's "Deep Researcher" template to write legal reports in seconds! It generates sub-questions, answers them, and compiles a report. Try it now.
🗺️ LlamaCloud & LlamaParse:
- Invoice Reconciliation Agent: Build an invoice reconciliation agent using LlamaIndex.TS and LlamaCloud to automatically check invoice compliance with contracts. This is a great real-world enterprise use-case. Watch a video of it in action or check out the full code here.
- LlamaParse for Financial Applications: Join our workshop hosted by CEO Jerry Liu to learn how to apply AI to financial challenges using LlamaParse. Register today.
✨ Framework:
- LlamaDeploy Updates: LlamaDeploy now supports a new message broker, Solace, enhancing its async-first framework for deploying agentic multi-service systems. Read more.
- New Model Alert: Qwen 3 is out, and benchmarks are impressive! Our friends at Ollama have already released support for it in LlamaIndex. Learn how to use it.
✍️ Community:
- Case Study on CondoScan: Discover how CondoScan uses LlamaIndex's workflows to create a next-generation condo evaluation tool, reducing document review time from weeks to minutes. Read the case study.
- Our thoughts on effective agent design: We weighed in on the optimal design of agents with with our strategy of mixing the autonomous nature of agents with the deterministic nature of Workflows.
