San Antonio is on the I-35 corridor, which runs south to Laredo and on, across the border, to Monterrey, where I live and work. The city is known for the military and cybersecurity, healthcare and bioscience, financial services and insurance, vehicle manufacturing and tourism. I take those five in turn below, starting with the one where my published work is closest, and I say for each what a company tends to need from AI and which service of mine answers it.
Vehicle plants and their suppliers
A plant and the companies that supply it tend to ask for the same things: answers drawn from their own procedures, small tools a supervisor can use during a shift, and numbers that separate one cause from another. LeanManufacturing.mx is the published work here. I built it for a lean manufacturing practice in Monterrey that works with automotive, metalworking and food plants. It has open documentation, a maturity self-assessment, an OEE calculator, an assistant grounded in the documentation, and an MCP server so that agents read the same material people read. The services that correspond are internal tools with AI and custom AI tools.
Insurance and financial services
An insurer lives on policies and claims, and on being able to explain a decision afterwards. No case study on this site comes from insurance or banking. The nearest is an analytics platform for warranty management, where agreements and claims scattered across an operational database were brought into one database and one dashboard. What transfers is that order of work: the records are made consistent first, and only then does a model read them.
Healthcare and bioscience
Hospitals and laboratories hold records that very few people may see. What they tend to need is control more than cleverness: each tool an agent can call sits behind one entry, each call is authenticated, and each leaves evidence. The published work that transfers is an MCP gateway on Cloud Run built that way, and the service is MCP servers and integrations. Nothing in it is medical. Your clinicians and scientists decide what a right answer is, and I build the access and the tests around their decision.
Military and cybersecurity
No case study on this site comes from the military or from cybersecurity. What transfers is a habit those fields already have: nothing acts without a record. The guide on agent auditability lists what the record of an agent's action has to contain. A program with rules about who may handle its data, and from where, applies those rules to me as to anyone else.
Tourism
Hotels and attractions answer the same questions again and again and pass the unusual ones to a person. That is the shape of the system in my education case study: agents that answer, schedule and route to the right human team for a higher-education group with 700,000+ students. What I publish about it is the evaluation gates that decide whether a change ships, and the service is AI agent development.
Nearly the same clock
From early November to mid March San Antonio and Monterrey are on the same hour. From mid March to early November San Antonio is one hour ahead of Monterrey. A call in the morning and a review of code in the afternoon fit inside one shared day.
The first step
Tell me by email what the process is and who owns it. I reply in writing with a first piece of work small enough to judge on its own. A senior role and a contract are both possible, and both are remote: I am not in San Antonio.