EDGAR Analyst
A retrieval-augmented generation (RAG) research agent built on every SEC filing since 1993. Ask a question and get an answer cited to the page and exhibit it came from.
The problem
A 10-K runs 200 to 400 pages, and an analyst can spend a full day reading one. Funds and research desks cover hundreds of companies, each filing several times a year.
General-purpose chatbots don't solve it. They answer from memory, can't show where a number came from, and haven't read the filing that landed this morning.
What I built
EDGAR Analyst indexes the entire SEC EDGAR archive, more than 22 million filings, and answers from the filings themselves. Every answer is cited to the page and exhibit, so an analyst can check it in one click.
On top of retrieval sit the tools analysts use every day: line-item extraction to spreadsheets, word-level redlines between periods, peer comparisons, and alerts the moment a company files.
What it does
Research
Ask in plain English. Answers cite the page and exhibit they came from.
Extract
Pull line items, segments and footnotes, and export them to CSV, Excel or JSON.
Compare
Word-level redlines of disclosures across periods, scored for materiality.
Peer analysis
Side-by-side comparison of up to 12 companies.
Monitors
Alerts by email, Slack or webhook when a tracked company files.
Free EDGAR reader
A public filing reader anyone can use without an account.
By the numbers
Under the hood
- RAG
- Next.js
- Fastify
- Postgres
- XBRL
- SEC EDGAR
- AES-256 at rest
- TLS 1.3
Want a system like this working for your business?
I design, implement, and optimize agentic systems that generate revenue, automate operations, and scale businesses. Tell me what you're working on.









