I work on Data & AI products at MWS (MTS Web Services): natural-language analytics, data governance, and AI assistants for enterprise data.

The problem is easy to state and hard to solve: companies already have the data, but it's scattered across systems, described inconsistently, and protected by access rules you can't ignore. A good AI product has to work inside that reality.

I lead products from discovery through pilots to go-to-market: defining the user problem, shaping the roadmap, working with engineering, testing with enterprise users, and deciding what's good enough to ship.

Two products I work on:

• MWS Data Scout — a knowledge layer for enterprise data: table and column descriptions, a business glossary, relationships between entities, and sensitive/PII data labels.
• MWS Data Copilot — a natural-language analytics assistant: a business user asks a question, reviews the plan, confirms, and gets back SQL, a table, a chart, or an export — with permissions inherited and every action audited.

I write about the product work behind enterprise AI: how to make LLM systems useful, measurable, secure, and boring enough to run in production.