DATA FAQ RECEIPTS.
Every number has a source. Every source is public. Here's how the machine handles both.
KARL: “I show receipts or I say I don't know. This page is the receipts about the receipts.”
Where does the data come from?
Every record comes from free public sources — the same filings and disclosures anyone can get, organized and weaponized. The backbone: FEC campaign finance, SEC EDGAR corporate filings, Senate lobbying disclosures (LDA), USPTO patents, OSHA and DOL wage-and-hour records, EPA violations, CMS nursing home data, NLRB labor cases, CFPB consumer complaints, FDA and CPSC recalls — 45+ datasets and counting.
No scraped paywalls. No proprietary feeds. No data brokers. If the public can’t see it, neither do we — and if the public can see it, we make sure everyone does.
How often is the data updated?
Nightly. Lux runs after hours: ingesting fresh filings, re-scoring the Ghoul Index, and hunting for new connections across the datasets. Scheduled ingests cover sources on their own cadences — FEC filings as they land, OSHA and EPA updates as they publish.
If a record exists in a public filing, it’s headed to the machine. Freshness is the whole game.
How accurate is the data?
Every record is sourced and cited — you can click through to the filing it came from and check it yourself. When the machine draws a connection between records (Lux’s job), it’s labeled as AI-detected, not fact, so you know the difference between “the filing says” and “the model suspects.”
Public records have errors, and we’d rather show you the source than pretend they don’t. Verify before republication — receipts are meant to be checked, not worshipped.
Why do your company numbers sometimes differ from official PR?
Because we use filings, not press releases. A company’s earnings call says one thing; its SEC filing, EPA docket, or OSHA record says another. We publish the filed version.
PR is marketing. Filings are legal. When the two disagree, the filing wins — and we’ll show you which one it is.
Can I download bulk data?
Yes, on All Access ($349/mo) and Enterprise ($1,499/mo) via the API — bulk downloads come with the tier. The free tier and individual packs are query-based, not bulk.
Aggregates are free to explore on the site anytime. If you’re building something serious, the API FAQ has the pricing.
How do I report a data error?
Email mtcstw@gmail.com with the record, what’s wrong, and the source that proves it. Or complete a data bounty on the Bounties board — corrected data earns XP.
We’d rather be corrected by you than embarrassed by a CEO. Send it.
Last updated October 2026. New datasets are announced in the feed.