Turn Complex Financial Docs into LLM-ready Data with Jerry Liu and LlamaIndex
A hands-on workshop for engineers building VLM-powered OCR that works on real-world financial documents.
Most document pipelines fail quietly. They work on a handful of PDFs, PPTs, XLS — then break the moment you scale across complex financial data with tables, charts, complex layouts, and more. Edge cases no amount of templating will ever account for.
FinParse NYC is a hands-on workshop hosted by Jerry Liu, CEO of LlamaIndex, where engineers take real, intentionally difficult financial documents and build a pipeline that helps them transform it into structured, usable data. We explore OSS alternatives as well as the value of using a commercial document ingestion API like LlamaParse that provides industry-leading accuracy, cost-efficiency and scale, as well as proprietary confidence scores and citations for HITL review.
This isn't a demo. You'll build a working pipeline on documents that are known to break things.
Agenda:
3:00-3:30 - Jerry Liu, CEO of LlamaIndex, presents on the future of agentic document OCR and how AI agents have transformed a trillion dollar legacy industry
3:30 - 5:00 - Hands on builders workshop
5:00 - 6:00 - Lessons from the field: How a Top 5 private equity firm built a risk decision making agent
6:00 PM - Dinner and Networking
What You'll Walk Away With
A working parsing + extraction pipeline you can extend
Practical schema design patterns for real-world financial documents
A clear understanding of why most document AI systems fail in production and what to do about it.
Location in midtown to be disclosed to all registrants shortly.