About
AI integration architect
Solution architect with 12+ years designing systems for regulated industries.
AWS 3× certified. Former solution review board member at the Ministry of Health.
Now focused on AI integration: moving LLMs past demos into production systems that hold up under audit, latency, and real traffic.
What I work on
Moving LLMs past the demo. Eval pipelines, failure-mode budgets, audit trails, and the plumbing that lets a model be a component of a system rather than the system itself.
FHIR, HL7, hospital-to-cloud integration. What it means to architect for an environment where every request might be the one an auditor picks.
Payments, settlement, compliance-adjacent workloads on AWS. When the SLA is tight and the regulator is real, architecture choices narrow fast.
Credentials
Let's talk
I work with a small number of teams at a time on AI integration architecture — usually healthcare IT or fintech — where the stakes are regulated, the latency budget is narrow, and the demo isn't the hard part.
Advisory
1–2 weeks
I read your existing design, interview your engineers, and hand back a written review: the failure modes I would worry about, the first three changes I would make, and the parts that are already in good shape. Output is a document, not a deck.
Build
4–12 weeks
I design the integration, write the reference implementation, and leave your team with a system they can extend — eval harnesses, observability wiring, and the documentation for the parts that aren’t obvious from the code.
Retainer
Rolling · 4 hours/month
For teams already mid-flight. A fixed block of my time each month for design reviews, reading diffs, or being the second pair of eyes on hard architecture calls. Small, reliable, cancellable.
Out of scope
Get in touch
Email deva@soa.team with a paragraph about what you're building and where you're stuck. I'll reply within a few days if I think I can help — and tell you honestly if I can't.
Or reach me on LinkedIn.