About

I'm a technology storyteller, a builder enabled by AI, and an AI research product manager at Meta.

My career took me through the trenches of some of the biggest platforms in tech.
- At AWS, I was the PM for the managed database portfolio, Elasticsearch, DynamoDB, RDS, and ElastiCache, learning the unit economics of cloud services from the inside.
- At Twitter, I led the effort to migrate AI and data workloads to Google Cloud and negotiated a ~$1B GCP contract.
- At Block, I helped the company to build the data flywheel.
- At Intuit, I built agentic AI workflows on their GenOS platform.
- Now at Meta, I lead AI research product strategy for the RecSys ranking models behind Instagram and Facebook recommendations.

Each of these roles taught me to see technology not as abstract systems, but as stories: about trade-offs, incentives, and the humans navigating them.

Today, AI has also made me a different kind of builder. I ship full-stack applications - like this very site and a bunch of others - using AI as a force multiplier. The line between product manager and engineer is blurring, and I'm thrilled for it.

What I write about

AI & Recommender Systems — Ranking models, agentic architectures, and where AI product strategy is actually headed (vs. where the hype says it is).

Building with AI — What it looks like to ship full-stack products as a PM-turned-builder, using AI to collapse the gap between idea and implementation.

Cloud Stuff — The hidden math behind managed services, GTM for various database variety, and enterprise cloud negotiations. Real numbers from real experience.

The name

I go by magicmag online.  yes, inspired by Magic Mike. Because why not bring a little showmanship to the world of AI and cloud? Great products feel like magic to their users, even when, especially when, the underlying systems are anything but. That's the story I'm here to tell. 😎