Spencer Varadi

I'm a technical program manager who likes the hard middle of a program: several teams, real dependencies, and a date that doesn't move.

At Curotec I ran product and delivery across 6 or more concurrent client engagements, from discovery through launch, and every one landed on time with 95% or better client satisfaction. Along the way I drove a company-wide Kanban transformation that cut cycle time 25%, lifted on-time delivery 30% and dropped defect rates 20%. I also built the reporting that roadmap decisions rested on, including a Metabase KPI dashboard drawing on about 20 source systems and a Postgres data warehouse with a 40-table ERD, and then deployed HubSpot and configured Jira and Confluence, which halved sales response time and added $500K+ to the revenue pipeline.

Comcast was the harder one. I was Product Owner for replacing a 20-year-old network design platform, with $100MM in field hardware at risk across 10 million miles of plant. I led three scrum teams totalling 30 or more people and held 98% sprint goal attainment. The platform shipped to 500 or more network designers across 13 regions. Throughput went up 35%. I wrote the user stories in Gherkin and defined done and ready at Comcast before carrying that practice into Curotec.

I also build with AI: an open-source portfolio control plane for governing autonomous software development, and an evidence-grounded research and verification platform on a PostgreSQL knowledge graph. That's why I write requirements with acceptance criteria, and why I'd rather settle ambiguity before a sprint than after a release.

If you're running a program where three teams owe each other dependencies and the date is already fixed, that's the work I want.