I’m starting this blog to write about the part of enterprise AI that rarely makes it into demos: turning messy company data into answers people can trust.

I’m a Product Owner for Data & AI at MWS. Day to day I work on natural-language analytics, data governance, and AI assistants for enterprise data — and most of the hard, interesting work is not the model. It’s context, permissions, evaluation, and the product decisions that decide whether a system actually gets used.

Posts here will be concrete and honest: what works, what doesn’t, and why. More soon.