wbaik/sapient.finance — explained in plain English
Analysis updated 2026-05-18
Browse AI-generated stock valuations ranked by base case upside for over 250 tickers.
Read cited research briefs behind each valuation's revenue growth, margin, and multiple assumptions.
Study a reference snapshot of a Next.js and Convex app that renders agent-generated financial research.
| wbaik/sapient.finance | 4ssh1/mine-sweeper | achawla19/intuitcode-extension | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | TypeScript | TypeScript | TypeScript |
| Last pushed | — | 2026-01-31 | — |
| Maintenance | — | Maintained | — |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 3/5 | 2/5 | 4/5 |
| Audience | developer | vibe coder | developer |
Figures from each repo's GitHub metadata at analysis time.
This is a read-only reference snapshot, the data pipeline and valuation engine live in separate private repos and are not included.
Sapient Finance is the public website for a finance publication that is entirely written by AI agents rather than human analysts. Large language models research individual US stocks, argue over the assumptions behind their value, and publish full valuations, with no person editing or writing the final text. Every stock the site covers goes through the same four step process. First, the AI agents write research briefs that argue for specific numbers on three key drivers of a company's value: how fast revenue will grow, what operating margin the business will settle at once it matures, and what valuation multiple investors should apply on exit. Each of those three numbers gets a pessimistic, a middle, and an optimistic estimate, and every estimate points back to the specific brief that argued for it. Those three numbers then feed into a fixed, rule based calculation engine, not another AI, which turns them into a fair value range for the stock. A separate mode handles companies growing so fast that their numbers need to be normalized before the calculation makes sense. Every number shown on the site comes directly from that calculation engine rather than being typed by hand or pulled loosely from the written analysis, and the site ranks more than 250 stocks by how much upside the base case assumptions imply. This particular repository is only the part of the system that renders the public website, built with Next.js and a backend service called Convex. It displays the research briefs, filing summaries, valuation packages, rankings, information about large fund holdings, and transcripts of the AI agent sessions that produced the analysis. The actual pipeline that gathers financial data, runs the research, and calculates valuations lives in separate, private repositories and simply feeds its results into this app, so this repository is a reference snapshot of the display layer rather than the full working system behind sapient.finance. The project describes itself plainly as an experimental system and states clearly that its output is research from an AI agent, not investment advice, and is not a recommendation to buy or sell anything.
The public Next.js and Convex website that renders AI-agent-written stock research and valuations for sapient.finance.
Mainly TypeScript. The stack also includes TypeScript, Next.js, Convex.
Setup difficulty is rated hard, with roughly 1day+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.