I run two related practices: shipping production agentic AI on Amazon Bedrock, and the AWS foundations that make any of it work. Every engagement is scoped, fixed-price where possible, and built in your account.
Feasibility, build, and the architectural muscle to make GenAI production-grade. The work that's the focus of the practice today.
3 weeks · Fixed price · Your AWS account
Three weeks. One use case. A working proof of concept in your AWS account, a real cost model, and a written go/no-go you can take to leadership. No retainer, no commitment to the next phase.
You've got an agent running. I review the architecture for reliability, cost, evaluation, and guardrails, then come back with a prioritized list of changes. Useful before a production launch, or after a near-miss.
03The next step after the Feasibility Sprint says go. A build engagement scoped from the Sprint's findings, taking a PoC to a production-ready service in your account. Always scoped separately, never assumed.
Patterns built up since the S3 days, and a lot of production AWS behind them. The infrastructure layer underneath everything else, and useful in its own right.
A secure AWS foundation deployed via CloudFormation. Multi-account setup, AWS Organizations + SCPs, IAM following AWS best practices, CloudTrail, GuardDuty, monitoring, and a reference architecture of your choice (Windows, Linux, or WordPress). Templates are yours.
05Next-gen AWS managed services. A certified AWS expert who designs, builds, and manages AWS alongside your existing team, carrying the load where you need it and handing back where you don't, at a fraction of the cost of hiring that depth in-house.
06A point-in-time review of your AWS environment against the patterns I've built up over years of running and auditing AWS accounts: IAM, networking, logging, data, edge, and incident readiness. You get a prioritized fix list and a report, walked through with me.
You send what you're trying to do. We talk for half an hour to see if it's a fit. If it isn't, I'll say so, and where possible point you somewhere better.
For Sprints and audits it says exactly what I'll do. For builds it says where the work stops. Either way, fixed price wherever the work allows.
Code into your repo, infra into your AWS account. Weekly checkpoints with a written status note. No surprises at the end.
You own everything. If you want a next engagement, we scope it. If not, that's the right outcome. The goal is your team owning what we built.
Most engagements start with a few sentences in an email. From there I'll either propose a Sprint, a different service, or, if it's not a fit, a better-suited recommendation.
