Document Analyzer
A team workspace where a company asks a grounded AI agent about its own policies, and turns spreadsheets into instant dashboards whose numbers are computed from the files.
OngoingStarted October 2026
What it is
Document Analyzer is a multi-tenant SaaS for companies. A team uploads its policies and handbooks, and members ask a grounded AI agent what the company does, may and must do. Members can also upload spreadsheets and documents and get an instant analysis.
Three tools
- Chat. Answers come from the company's own documents, through hierarchical RAG partitioned per team.
- Dashboard. Text cards and charts built from uploaded files, with a sidebar chat that answers and edits the dashboard. The AI plans the chart and the server computes the numbers, so a figure on screen always comes from the file.
- Advisor. Most small companies have no management accountant. A member uploads sales, costs and stock, and gets a report with up to seven concrete recommendations, each naming a product and its estimated monthly impact.
Teams and billing
Accounts sign in with a phone number and a password and can belong to many teams, with admin, manager and staff roles per team. A team's first 50 credits are free, and after that it buys credit packs with QPay or Stripe. Credits meter the AI cost.
How it is built
A Turborepo monorepo with a NestJS GraphQL API, two TanStack Start apps and a Python AI service: FastAPI and an arq worker running LangChain and LangGraph with Claude Haiku, Voyage embeddings and rerank over pgvector. It has not launched yet; the screenshots show development data.
