FN2

FN2 is an autonomous AI research workspace for public markets: build an agent once and it runs each market day on live Polygon prices, SEC EDGAR filings, FRED macro data and Polymarket odds, returning briefings where every claim carries a citation.

FN2 is an autonomous research workspace for public market investors. Rather than answering one question at a time, it lets you build a research agent once and then runs it every market day against live data. Outputs arrive as scheduled briefings: SEC filings summarized, earnings call metrics pulled, macro series read, and every claim tied to a numbered citation so you can check the underlying document. The product leans on four public data sources. Price quotes, bars and charts come from Polygon, filings are read directly from SEC EDGAR, macro context is drawn from FRED with year over year and month over month figures on each series, and prediction market odds from Polymarket sit beside the price action they explain. In practice you ask something like what drove a five percent move in a ticker, and get a written answer with the catalyst, the relevant quotes, a sector comparison table and a citation on each line. Separate live demos show chart lookups, earnings analysis, macro reads, market screening, transcript search by meaning and prediction market pricing. It suits self-directed investors, analysts and small research desks that want institutional-style coverage without an institutional terminal. The founder frames the product explicitly against Bloomberg-style terminals and trading desk tools. Because the sources are public and verifiable, answers can be audited rather than taken on trust, which matters more for filings and earnings work than for casual chart browsing. A browsable earnings page gives every reporting company the same cited data without going through chat at all. For technical users there is an open API and plugins for external agent frameworks such as Hermes Agent and OpenClaw, so a self-built agent can pull grounded market research on demand or create, schedule and run FN2 loops itself. That makes FN2 a finance layer under an existing agent workflow as much as a standalone app. It is not financial advice and it is not a brokerage or portfolio system.