Sometimes you need to start over. But that decision is hard.

@llama_index had to make that call: they built one of the most popular AI frameworks in the world, but saw the agent harness and frontier labs coming for them.

So they drilled down on a durable moat: extracting enterprise data, and providing the best, most accurate context.

@jerryjliu0 joined me on the @chain_ofthought to talk about how they iterated, why they disrupted their own product before a frontier lab could, and how LlamaIndex became the best in the world at turning messy PDFs and PowerPoints into context enterprises can actually leverage for their agentic workflows.

As Jerry put it: pointing Opus 4.7 or GPT 5.5 at your document corpus feels like 80% accuracy in a demo, but that's a deception. The other 20% hallucinates a number or misreads a table, and your whole agent workflow breaks. In legal, insurance, and finance the real bar is 95% plus.

That's what LlamaIndex is focused on today - watch the full, frank conversation with Jerry below👇

Chapters: 0:00 Is the AI framework era over? 1:56 What died and what survived 6:31 Why context quality is the moat 8:12 Defining the context layer 13:18 Coding and vision as the abstraction layer 18:13 The bet that context compounds 23:59 Which verticals are adopting 25:14 Why 95%+ accuracy is the real bar 29:49 The file system as an agent primitive 34:33 Surviving your own pivot 37:15 Reinventing strategy and hiring 42:00 Agent memory as persistent context 44:41 Model personalities and cultural memory 47:51 Writing with AI 50:19 Closing thoughts