GAPIVO
$70,000
the debt one person was chased for - a debt they say was never theirs.
We read 100 real complaints people filed with the US consumer finance regulator (the CFPB) in one recent month. Names and personal details were removed before we saw them. Everything on this page traces back to those complaints - and everything we could not confirm is marked as a guess.
2 of 2 ideas on the board have a payer we can name.
Software that reads a consumer's complaint and tells a lawyer, in minutes, whether there is a real legal case in it.
The risk that kills it: Attorneys do not trust the triage - one hallucinated element or missed statute of limitations in a demo and the product is dead in that firm forever. The mitigation (citations back to the client's own words for every extracted element) is also the product's core design constraint.
A watchdog service that tells a bank or lender what patterns are forming in public complaints about them - before those patterns become lawsuits.
The risk that kills it: The buyer already gets this for free: companies receive their own CFPB complaints directly from the portal, and if internal complaint data is a superset of the public feed, the product is a nice-to-have dashboard. The counter-hypothesis to test first: internal teams see trees, not forest - the value is clustering and early pattern detection, not access.
Build plans, cost models and weekly updates arrive with membership. The ideas, their evidence and their honest risks stay free.
Everything we read was US banking and lending. A different set of complaints might point somewhere completely different. And 100 is a small sample - treat every number here as a starting point, not a proof.