Who's behind MouseCat?
Nick — one of the core maintainers of MCP, the protocol AI agents use to talk to tools.
Joe — built risk infrastructure at Coinbase.
Agents + fraud plumbing, from the people who built both.
If your team is drowning in flagged cases, let's talk
Writing fraud rules usually looks like: notice a pattern, write a rule, ship it, hope it doesn't block good customers.
MouseCat flips it:
→ agents spot the pattern across investigated cases
→ build features from your warehouse, point-in-time.
The slowest part of fighting fraud: you often don't know it was fraud until the chargeback shows up weeks later.
By then, the same pattern has hit you 200 more times.
MouseCat generates labels for account takeovers and chargebacks right after investigation .
Most fraud tools do one thing: score in real time, or help humans work cases.
MouseCat connects both:
Investigate → every flagged case, fully
Learn → patterns across what agents found
Prevent → new rules, backtested before they go live
The math of a fraud queue:
100 cases flagged.
5 get a real investigation.
95 get decided by a score nobody can explain.
That's where losses — and angry legit customers — come from.
MouseCat's agents work all 100.
Most fraud teams review ~5% of flagged cases.
The other 95% get a guess.
Meet MouseCat — AI agents that investigate every single case like a senior analyst would: pulling history, checking the merchant, tracing linked accounts, even calling the phone number.
99K Followers 5K Following#Fan Account• From The Forest To The Chain•Hunting Memes & Gems On #Robinhood chain •DM For Collabs 📩 No Affiliation @Robinhoodapp
50K Followers 2K FollowingHunting Gems Before They Pump 💎
BTC | ETH | SOL | BNB
Daily Alpha • Early Calls • Web3 Insights
📩 DM For Business / Promo @PoolKinggg
62K Followers 2K FollowingWe RH Tribe 🏹 • The leading community for Robinhood Chain memes • Curating the hottest plays on RH • Always DYOR • Not affiliated with @RobinhoodApp