← back
RAG for VPs of AI: Jerry Liu
Takeaway
Enterprise RAG success hinges on a dedicated data-processing stack and a bet on in-house developers, with parsing quality (e.g., LlamaParse) being the single biggest lever against hallucinations.
Summary
- Jerry Liu (CEO LlamaIndex) frames enterprise RAG as needing a new data-processing stack distinct from analytics ETL — PDFs sliced, indexed and stored across vector, doc, graph or SQL stores so LLMs can synthesize answers.
- Prototype-to-production is the central pain: a 10-minute RAG pipeline that 'kind of works' degrades when document count, complexity and source variety grow; tuning knobs and data silos kill ROI for big-up POCs.
- Argues VPs of AI should bet on developers and custom builds over out-of-the-box tools because the technology shifts too fast for procurement-led solutions.
- LlamaIndex offering: open-source toolkit for orchestration of retrieval/prompting/agentic reasoning + LlamaCloud as a centralized knowledge interface unifying data sources and enhancing quality.
- LlamaParse (advanced doc parser) processes complex financial reports, PowerPoints with messy tables/diagrams; cites ~500k monthly client-SDK downloads and tens of millions of pages parsed.
ragllamaindexenterprise
Original description
Ask anything about the state of Retrieval Augmented Generation with the CEO of LLamaIndex. 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 Llama Index