← back
The Future of Knowledge Assistants: Jerry Liu
Takeaway
Move from naive RAG to agentic, multi-agent knowledge assistants built on high-quality parsing and tool-using LLM orchestration.
Summary
- LlamaIndex CEO Jerry Liu argues naive RAG is just a glorified search system and outlines three layers to evolve into a real knowledge assistant: advanced data/retrieval, agentic single-agent flows, and multi-agent task solvers.
- Good PDF parsing (e.g., LlamaParse) materially reduces hallucinations versus PyPDF; demos a CalTrain schedule and financial-table example where structured parsing fixes wrong answers.
- Adds agentic layers — function calling, query planning (sequential, DAG, tree), memory, ReAct loops — turning LLMs into orchestrators over data services as tools.
- Single agents fail with thousands of tools and stateless designs; recommends specialist agents and a multi-agent future for knowledge tasks.
- Announces LlamaCloud productization for enterprise PDF processing at scale (tens of thousands of users, tens of millions of pages).
ragagentsllamaindex
Original description
In this talk, LlamaIndex founder & CEO Jerry Liu covers how we go beyond single-LLM prompt calls. He discusses advanced single-agent flows, Agentic RAG, multi-agent task-solvers & service architectures, and more. Jerry also announces Llama Agents: Agents as microservices that are easily deployed and communicate via a single API (and much more). Recorded live in San Francisco at the AI Engineer World's Fair. See the full schedule of talks at https://www.ai.engineer/worldsfair/2024/schedule & join us at the AI Engineer World's Fair in 2025! Get your tickets today at https://ai.engineer/2025 About Jerry Jerry Liu is the CEO & Co-Founder of LlamaIndex