Why it matters
Several competing RAG/agent-tooling stacks get demoed back-to-back in one room, letting practitioners see real ingestion and retrieval tradeoffs side by side instead of reading vendor docs in isolation.
The tokenmaxxing angle
OpenRAG and Unstructured.io take different approaches to document ingestion and chunking, which directly changes how many tokens a RAG pipeline burns per query — a rare chance to compare their retrieval overhead live.
From the organizers
Agenda lists tech showcases from OpenRAG, Unstructured.io, IBM Bob, and PyData at 75 Binney St, Cambridge, 5:30-8:30pm, hosted by IBM via the 'Data, Cloud and AI in Boston' group.