From RAG to Riches, revisited: a case study with Polygrade

Isaac Arnold of 11 Tribes Ventures published “From RAG to Riches, Revisited,” a case study that uses Polygrade to test an argument he made earlier this year: in AI, owning the model is rarely the moat. Durable companies win by building domain expertise into the judgment layer around the model, and that advantage is gated by founder-market fit.

Written with contributions from founder and CEO David Steckel, the piece grades Polygrade against five moats: workflow entanglement, domain-specific abstraction layers, distribution and trust, human-in-the-loop learning, and cost and margin. It also walks through how Polygrade takes risk off the buyer: a claims analysis on a prospect’s own claims, time and motion studies, running on top of existing systems, and implementing one module at a time.

11 Tribes Ventures has not invested in Polygrade. The piece grew out of an ongoing conversation between Isaac and David about how AI is changing the market.