A RAG learning product needs a sales model
A partly built RAG product, about 60% complete, aims to help people learn faster and make better decisions. It has about 30,000 chunks of content prepared for search and answers.
Users can ask questions and get quick answers instead of digging through lots of material. If they forget something and ask again later, the system is expected to return the same answer.
So far, the experience has been that answers come without . The open question is how to package and sell it: a with prices, a purchase, a , or local installation on the customer’s own computer.
Key points
- The product uses RAG to answer questions from prepared learning content.
- About 30,000 chunks of content have already been created.
- The goal is fast, repeatable answers with less need to reread source material.
- The main un choice is packaging: purchase, , or local install.