Journey
Started tinkering with PC builds and web projects in college. Moved into AWS, DevOps, and full-stack development at Via.com. Crossed into product at ZipGo, SIXT, Hevo, and ThoughtSpot. Now building AI products at enterprise scale — and shipping my own on the side.
Full work history →Day job
What I do full-time.
Enterprise AI · Principal PM
Spotter @ ThoughtSpot
Natural language search and agentic analytics for enterprises — ask questions of your data in plain English and get governed, trustworthy answers. I worked on and launched Spotter Analysts: every team gets its own AI analyst. Pick the certified data models, tools and instructions, and a Supply Chain Analyst answers in supply chain terms.
See Spotter →Side builds
Things I built and run myself.
Travel blog · MCP-connected
WanderingBong
A slow-travel blog I've run since 2015 — through WordPress, Strapi, and eventually a backend I built myself. Now MCP-connected: agents handle tags, linking, SEO, and metadata. I show up with the experience. They handle everything around it.

Experiments
Built to learn. Some still running.
Knowledge Graph · MCP
Eidetic
Chrome extension + mobile app + MCP server. Captures anything I read, makes it searchable by Claude mid-conversation.
Personal OS · MCP
Jarvis
An agent-facing layer over my own workspace — context and tools an agent can act on directly, like giving it a butler's briefing instead of a chat window.
NLS · Benchmarking · Open source
BIRD → Snowflake
Nobody had built a Snowflake loader for the BIRD NL-to-SQL benchmark. I needed one for evaluating Spotter, so I built it and open-sourced it.
Data Viz · 3D Globe
Worldview
Every tracked satellite in orbit, right now. CelesTrak TLE data, satellite.js propagation, Mapbox GL v3 globe. Scrub through time.
Earlier
Where it started.
Before product, I was a developer. Built SEO systems and frontend tooling at Via.com — a Node.js + Redis content platform across 8 countries and 5 languages, driving a 200% increase in organic traffic. Before that, AWS, DevOps, and more PC builds than I should have spent money on.
Full work history →
