E134: Making Complex Data RAG-Ready with Unstructured

Brian Raymond is Founder & CEO of Unstructured, the platform to extract and transform complex data for use with every major vector database and LLM framework. Their open source project has 7K stars on GitHub and includes libraries and APIs that let users build custom preprocessing pipelines for labeling, training, and production machine learning pipelines. Today, they have over 6M downloads and 50K companies using their tools. Unstructured has raised $65M from investors including Bain, Essence VC, and Menlo Ventures. In this episode, we dig into Brian's process of talking to 100 data scientists before launching Unstructured, why the long tail of data matters for LLMs, competing with their own open source, why being a "boring company" is valuable for today's LLM stack, why they liked having government design partners, why world-class design & marketing are huge differentiators for open source companies & more!