Thirty years of shipping systems that can’t fail.
I’ve spent my career as the specialist organizations call when the migration has to land on time, the system has to stay up under real load, and the work can’t be learned on the job.
My current focus is production AI and data platform engineering in regulated healthcare — agentic systems, semantic document intelligence, petabyte-scale Snowflake platforms, near-real-time ingestion. This is the work most interesting to me right now and where the economics of specialist expertise are most pronounced: the clinical, legal, and operational constraints are real, and systems have to actually work on day one, not after a demo period.
Before this I spent three decades in mission-critical infrastructure. I authored the internal migration runbook for Amazon’s original move to AWS, back when the managed-service ecosystem most engineers take for granted didn’t exist and everything was built from EC2, S3, and EBS primitives. I architected the server orchestration abstraction layer that manages 110,000 Oracle Fusion SaaS servers across public cloud, FedRAMP, and private cloud — and co-invented a U.S. patent for electronic voting security along the way. I led the migration of New Jersey’s voter registration system onto a 13-site multi-master topology that has held 100% uptime through hundreds of election cycles. I served as primary technical advisor for Fortune 100 and government accounts at MariaDB. I built the MySQL automation framework that became the foundation of HP’s Database-as-a-Service public cloud offering.
The thread across all of this, at least as I see it, is an instinct for the work that matters structurally rather than cosmetically. The platforms I’ve built tend to outlast the engagements that produced them. The New Jersey election system is still running the elections it was built for. The AWS migration runbook became how an entire organization moved. The HP automation framework became a product. I like that kind of work — the kind where someone will still be using what you built ten years from now, and doing it without thinking about you.
How I engage
I take a small number of engagements per year, typically ranging from three-month technical transformations to multi-year advisory relationships. The work I’m most interested in sits at the intersection of data infrastructure, production AI, and regulated-environment constraints — places where both the technology and the stakes are real.
Most of my inbound comes through referrals, past colleagues, and people who’ve seen the work. If you’re facing a data platform, AI, or infrastructure problem that your team would need a year to ramp up on, and you’d rather just hire someone who’s already done it, I’m probably findable through the usual channels. adam@awre.io is the fastest path.
Outside of client work I still ship things for the fun of it — most recently somethingweird.today, an agentically-curated guide to genuinely weird date ideas, built on the same production-AI pipeline instincts as the day job.