AI Engineer
← All organizations

AI document processing and agent development

LlamaIndex

LlamaIndex builds document processing infrastructure for developers and enterprises automating knowledge work. LlamaParse turns complex documents into machine-readable text and structured data, with extraction, classification, splitting and searchable indexing. Its pipeline combines OCR, computer vision for layouts and LLM reasoning for tables and charts. LlamaCloud connects, parses, extracts and indexes enterprise data, while LlamaAgents lets users build document workflows through code or plain-English descriptions. These tools support document research, report generation and other workflows over proprietary information.

Founded in 2023 by CEO Jerry Liu and CTO Simon Suo, the company grew from an open-source framework launched in late 2022. Its indexing, retrieval and query-engine abstractions helped developers connect data to LLM applications. In 2026, LlamaIndex sharpened its focus on document infrastructure beyond RAG frameworks. Its open-source LiteParse CLI and TypeScript-native library runs locally, preserving spatial relationships by projecting text onto a grid. It complements LlamaParse’s cloud service with text, screenshots and bounding boxes for agents.

As of August 2026, the company reported 300,000+ LlamaParse users and 25 million+ monthly package downloads. It closed a $19 million Series A led by Norwest Venture Partners in 2025, bringing funding to $27.5 million at that point. Later that year, Databricks and KPMG made minority equity investments.

www.llamaindex.ai ↗

5 talks

Newest first

2 speakers at AIE

Affiliations reflect their AIE appearances, not necessarily current employment.

Jerry LiuCEOLaurie VossHead of Developer Relations

Start here

  1. Building Production-Ready RAG Applications

    The essay based on Jerry Liu's talk explains small-to-big retrieval: finding precise evidence in smaller units before supplying broader context for answer synthesis.

    Jerry LiuAI Engineer Summit 2023

  2. The Future of Knowledge Assistants

    The essay based on Jerry Liu's talk traces how an assistant can plan queries, choose services, and combine their outputs instead of sending every request to the same vector database.

    Jerry LiuAI Engineer World's Fair 2024

  3. Effective agent design patterns in production

    Laurie Voss's talk provides a starting point for deciding when unstructured data and software-integrated workflows warrant an agent rather than a simpler RAG or chatbot approach.

    Laurie VossAI Engineer World's Fair 2025

  4. Building AI Agents that actually automate Knowledge Work

    LlamaIndex co-founder and CEO Jerry Liu explains how AI agents can automate document-centric knowledge work by moving beyond basic RAG toward document toolboxes that integrate enterprise data sources, permissions, indexing, search, and document manipulation.

    Jerry LiuAI Engineer World's Fair 2025

Messages from the stage

Measure retrieval before adding complexity

Liu distinguishes complete response evaluation from retrieval-specific testing. Benchmarks provide a basis for choosing improvements such as chunk-size tuning, hybrid search, and metadata filters.

Document structure shapes agent capabilities

Liu connects structured parsing and reliable ingestion to more capable knowledge assistants. His document automation talk extends that foundation to spreadsheet normalization, document manipulation, and human oversight.

Give agents bounded responsibilities

Liu describes specialized agents with smaller tool sets and independently orchestrated services. Voss adds production patterns for routing, parallelization, and aggregating outputs from multiple model executions.

Affiliations reflect each recorded session, not necessarily current employment.

Company sources · checked 2026-08-27