I lead product for regulated financial platforms, and outside of that I build and ship AI tools solo. Nothing below is a claim you have to take on faith. Each one has a link, a number, or a name attached to it.
Every portfolio makes claims. Here's what backs up mine.
What I've actually gotten good at, and where it came from.
Own the backlog end to end and run it through Marketing, Sales, Customer Care, Finance, Engineering, and Security/Risk stakeholders, not just Engineering. A-CSPO and CSPO certified.
Build a working version of an AI feature before writing the spec for it. That's what produced FitCheck and DharaPutra, both shipped solo, and a local multi-agent Ollama setup running on my own machine.
Treat regulatory constraints as non-negotiable inputs to the spec, not an afterthought. FitCheck was built to DPDP 2023 requirements from day one: consent flow, data retention, cross-border disclosure. Not retrofitted later.
Built and led a 10-person product team from zero at R Systems. Managed and mentored a 7-person BA team at Conduent as Lead BA.
Write acceptance criteria as the actual verification layer, not documentation theatre, and build in a skeptical second pass before calling anything done. Same standard applies to my own CV: no fabricated metrics, ever.
Know enough about the trade-offs to make a real architecture call, not just approve one. DharaPutra went through a full pivot from Flask/Redis/Render to Cloudflare Workers/KV once the edge-native version proved out faster and cheaper.