Making unfamiliar documents reliable.
Trust mechanism: evaluation before release, humans wherever confidence is low.
100K+documents/month
Read case studyI build AI that enterprises trust in production.
Technical PM across AI platforms and enterprise SaaS.
I’ve shipped evaluation-led document AI at 100K+ docs a month, founded an inventory platform that exited through a technology sale, and launched a high dosage tutoring platform in the US.
95%+ field-level exact match before human review
ProductSquads63paying customersFounder, exited through a technology/IP sale
Closphere$1.5MARR in six monthsUS marketplace launch, 120K users
FiloTrust mechanism: evaluation before release, humans wherever confidence is low.
100K+documents/month
Read case studyTrust mechanism: continuous reconciliation against the source ERP, plus silent-failure detection.
63paying customers
Read case studyTrust mechanism: confidence tiers, price floors, and human override on every quote.
18h → <1hin-stock quotes
Read case studyTrust mechanism: a 60-second matching promise, kept for 98% of US requests in under 45.
$1.5MARR in six months
Read case studyMy path spans enterprise document AI, an inventory platform I founded, and a US tutoring marketplace launch. The thread is the same: make complex systems dependable enough for people to act on.
More about meUW MSIM, Product & AI, Sep 2025 to Aug 2026.
Amity BTech CSE, 2016 to 2020.

For senior PM and AI product opportunities, or a complex problem worth working on.