I own the agentic workflows behind Amazon Quick Suite, our generative AI cowork platform for enterprises. More than 10,000 Amazon employees now use the workflows I shipped. I run discovery straight with C-suite stakeholders, validate demand before we spend engineering, and treat proof-of-concept adoption as the signal for what earns a place on the roadmap. That discipline pushed prototype-to-production adoption to 70% and turned a stuck 30% portfolio win rate into 45% over two quarters.
I build AI products that people actually adopt.
I am a Senior Product Manager at AWS working on generative AI for enterprises. For eight years I have turned ambitious ideas into products that ship, get used, and move real numbers. The part I care about is not the launch. It is whether anyone comes back the next day.
Three products, one habit: prove it gets used before we scale it.
A closer look at the work behind the numbers, in my own words.
I led Mission Cloud's first SaaS product from zero to one. It gave AWS enterprise customers real-time visibility into their cloud spend and automated recommendations to act on it. The optimization engine, Mission Cloud Score, cut customer AWS costs by 22% on average and lifted CSAT to a 92% rolling average. The proactive features reversed an 8% churn trend and kept 80% of at-risk customers, which is the part I am proudest of, because it meant we had built something people relied on.
Earlier, at Amadeus, I built a hotel loyalty portal from scratch and rebuilt the guest management system behind it. Wiring it into booking systems and third-party apps lifted loyalty signups 22% globally. The $30M phased UX overhaul I led pushed customer retention from 74% to 89%, and a new drag-and-drop email tool raised open rates by 41%.
More projects on GitHub, added as I build them.
Three things I believe about building product.
Validate before you build
I would rather kill an idea in discovery than ship something no one asked for. Proof-of-concept adoption tells the truth long before a roadmap does.
Adoption is the real metric
A launched feature that goes unused is not a launch. I instrument everything and follow how people actually behave, not how we hoped they would.
Stay close to the craft
I build with the tools, not just about them. Working hands-on with Claude Code, MCP, and agents keeps my product judgment honest and current.
The through line in my work and my life is the same: patience, small adjustments, and paying attention to how things actually land.
I roast my own coffee beans, which turns out to be a full course in patience and tiny variables. A few seconds and a few degrees change everything. Product work rewards the same attention.
Most of my time outside work goes to my growing family, training, and following far too much sport. I grew up speaking Greek, studied computer and electrical engineering at Farmingdale State, and I have been building products in and around New York ever since.
If you are hiring a product leader for AI, I would like to hear about it.
The fastest way to get a feel for my work is the resume. The best way is a conversation. Either one is a click away.