John Moody · Richmond, Virginia
I work with founders who have a working prototype, or an AI-assisted codebase that works but nobody has hardened. I make it something you can run a business on: tested, monitored, and safe to deploy.
How an engagement worksThree systems I built and ran in production, all in indoor-farming IoT. I’ve also worked in autonomous vehicles and robotics, and at a biotech startup. The full case studies are on the work page.
Data infrastructure
There was no data warehouse, and operational history was quietly disappearing as records aged out of each source system. I built the first one, solo, and designed it to be maintained on a short daily review: schema drift is detected automatically, and each morning an AI agent opens a migration pull request with a written rationale for me to approve.
Data access
Data questions queued behind the few people who could write SQL, and many were never answered. I built an internal MCP data-access server so non-technical staff could ask in plain English, with a concept dictionary underneath it so the system knows the difference between a seed date and a germination date. The people asking the questions now answer them.
Automation
Customer sustainability reports were built manually from several tools, one at a time, by whoever had the specialist knowledge. I replaced that with an end-to-end pipeline: scheduled reports run without anyone touching them, and the customer experience team generates custom ones on their own.
Availability
I’m currently taking on one new engagement. If it might be yours, email me a couple of sentences about the project and where it stands.