Launched this week
Disclosure Alpha
Transform complex regulatory disclosure into structured data
5 followers
Transform complex regulatory disclosure into structured data
5 followers
Disclosure Alpha is an open-source Python tool that parses, scores, and diffs 10-K, 10-Q, and 8-K filings locally and deterministically. Stripping out expensive LLM dependencies, it extracts 10 native, headline-weighted language scores to flag immediate text shifts at zero cost. Built for pragmatic quantitative and developer workflows, it runs seamlessly using local HTML pipelines or connects instantly to your AI environment via an integrated MCP server.
Hey Product Hunt! 👋
I built Disclosure Alpha out of a simple frustration: processing raw SEC regulatory filings (10-Ks, 10-Qs, 8-Ks) usually requires burning endless API credits on LLMs or wrestling with heavy, over-engineered enterprise datasets just to extract basic textual shifts.
I wanted something lightweight, fast, and entirely deterministic.
What Disclosure Alpha does out of the box:
No LLMs Required: It parses, scores, and diffs corporate filings locally without relying on external APIs or risking hallucinations.
Granular Text Scoring: It computes 10 distinct language scores (including 9 headline-weighted metrics) so you can instantly pinpoint exactly where and how a company’s tone or disclosure details changed quarter-over-quarter.
Native MCP Support: It features built-in Model Context Protocol (MCP) server capabilities. If you use AI assistants or local agents, they can connect directly to parse and explore these disclosures programmatically.
The project is entirely open source and built for developers, quantitative analysts, or anyone trying to extract clean data from messy financial disclosures without the bloat.
Check out the project website here: https://disclosurealpha.com
I’d love to hear your thoughts, answer any questions about the scoring logic, or get feedback on what features you'd like to see next!