Open to new opportunities
Most recently Director, AI Center of Excellence at Mastercard — leading enterprise AI and data platforms at scale. Outside of that, I build and run SatornTCG, a trading-card resale business, using it as a proving ground for the same class of tools I work with professionally, including hands-on MCP and LLM tool-use.
18+ years scaling AI and data platforms in Fortune 500 environments — most recently leading global engineering teams at Mastercard supporting 2,000+ users on enterprise AI infrastructure, including Kubernetes-based ML platforms and internal LLM tooling. The projects below are where I get hands-on personally with current agentic-AI tooling, including the Model Context Protocol.
Led a global team of 11 engineers supporting 2,000+ users; migrated 6,000 notebooks across 1,000+ servers to Cloudera AI, and built an AI-powered support chatbot that cut ticket volume 30%.
Built the org's first enterprise open-source Python distribution (100+ libraries, PCI-compliant) and enabled fraud-detection analytics processing 143B+ transactions annually via graph database technologies.
Drove Hadoop modernization scaling beyond 2 petabytes and led enterprise reporting migration off Business Objects, cutting platform costs 33%.
Full history, including earlier roles at PepsiCo and education/certifications (MBA, BS Computer Science, MCP: Hands-On with Agentic AI), is in the resume.
Two pieces of the same system: the commerce platform the business runs on, and the tooling that lets an LLM query it directly.
The commerce platform behind SatornTCG: inventory, pricing alerts, eBay/TCGplayer listing management, and profit & loss tracking for a live trading-card resale operation, backed by Postgres.
An MCP server that exposes the same inventory, pricing, and P&L data as callable tools for LLM clients — so an AI assistant can answer real questions against live business data instead of a canned demo set.
Diagnosed and fixed a Windows TLS-interception issue that only surfaced when the server was launched by a GUI client rather than a terminal — documented in the README. Built as hands-on follow-through on MCP: Hands-On with Agentic AI.
Both projects are active, not archived.