When AI actually belongs in your product (and when it doesn't)
A pragmatic framework for deciding whether a feature needs machine learning at all, starting from the metric you're trying to move, not the model you want to use.
read more →I'm a product and R&D leader who connects technology, data, and business. Over 12+ years I've taken connected products from a blank page to real-world deployment, and once, all the way to acquisition.
My background is hybrid by design: I started in applied research and engineering (photovoltaics, connected health, national digital platforms), moved into technical architecture and product management, and co-founded a company where I owned R&D, product, and operations end to end. That mix means I can sit with engineers on system design in the morning and defend a roadmap to customers and investors in the afternoon.
I care most about the moment a messy, complex innovation becomes something a real user actually wants. Lately I'm focused on where AI and MLOps make products genuinely better, not because it's a trend, but where it moves a real metric.
A progression from applied research to product leadership, each step with more ownership than the last.

Built an AIoT + SaaS solution for predictive roof risk management, from concept to real-world deployment and acquisition in 2026.





What I reach for across product, data, and engineering.
Open to product and tech roles where AI, data, and connected systems meet real users. If that's you, my inbox is open.