Funnel forensics using AI
AI agents read code, logs, events and user records in parallel to do RCA of funnel breakages.
Cash loans, trading systems, government transactions. Four years across three industries where the model, the money and the regulator all have to agree.
I'm a product manager who builds - I write the decision logic, wire up the systems it depends on, and ship the screen the user sees.

Shipped, not specified.
AI agents read code, logs, events and user records in parallel to do RCA of funnel breakages.
A memory layer for AI coding tools: the decisions that live in people's heads, written where any tool can read them.
A multi-model orchestrator on top of Nexus, so each task goes to the model that is actually good at it.
PRDs and prototypes.
Laid down the inception with eligibility gating. Built into a recommendation system, with ML models powering decisioning and lifting user approvals.
OpenOwning a lender journey end to end: the screens, the vendors, the edge cases and the architecture behind it.
ReadA handful of calls and a link: compose a lender's APIs into generic ports and ship it as configuration.
ReadA lending app, a trading floor, and a ministry.
Consumer fintechBachatt App · Sole PM, reports to the CEO
GovernmentMinistry of Finance (DIPAM) · Advised the Joint Secretary
Quant financeD. E. Shaw
Things I worked on that made the news.
On Substack.
Decision logic, messy integrations, and what the data actually said.
Read on Substack →If the hard part of what you're building is the decision, I want to hear about it.