Nearshore AI engineer in Mexico for US teams

I am a Senior AI Engineer and I work remotely from Monterrey, Mexico, with teams in the United States. Nearshore here means one engineer, under my own name, inside your working day, who builds in your repositories and your cloud accounts.

A team in the United States that looks for a nearshore AI engineer usually wants three things at once: a senior engineer, a clock close to its own, and work that stays with the team when the engagement ends. This page says what each of those means when the engineer is in Monterrey, how such an arrangement is usually set up, what I can show, and what this is not.

What nearshore means in practice

The clock. Monterrey is on UTC-6 all year. The central zone of Mexico no longer changes its clocks and the United States does, so the difference moves twice a year:

  • Central Time (Austin, Dallas, Houston, San Antonio, Chicago): the same hour as Monterrey from early November to mid March. From mid March to early November those cities are one hour ahead.
  • Eastern Time (New York, Miami): one hour ahead of Monterrey from early November to mid March, and two hours ahead from mid March to early November.
  • Mountain Time (Denver): one hour behind Monterrey from early November to mid March, and the same hour from mid March to early November.
  • Arizona (Phoenix): one hour behind Monterrey all year, because Arizona does not change its clocks.
  • Pacific Time (San Francisco, Los Angeles, San Diego): two hours behind Monterrey from early November to mid March, and one hour behind from mid March to early November.

At its widest the difference is two hours, so most of a working day is shared with every one of these cities.

The languages. English is the working language with a team in the United States: documents, code review and calls. Spanish matters when part of the operation is in Mexico, or when the people who will use the system write in Spanish. The tests of an agent belong in the language of its users.

Where the work lives. The code is written in your repositories and runs in your cloud accounts, with the access you grant. When an engagement ends, the system and its history are already where your team works.

How it is usually arranged

Two forms are usual, and I describe them only in general terms. In the first, the engineer is a contractor: there is a contract for services and an invoice for the work. In the second, the engineer is an employee, either of an entity that your company has in Mexico or of a third party that employs people there on behalf of companies abroad. Which one fits is for your legal and tax advisers to say, not for this page.

Whatever the form, four things belong in writing before the first day: the scope and what counts as done, who owns the code, which systems and data I may reach and how that access ends, and the hours of overlap. For my part, I am open to a full-time senior role and to contract engagements.

What teams in the United States tend to ask for

A software company or a startup often has an agent that works in a demo and no rule for what may ship. The work there is LLM evaluation: cases kept next to the code and a gate in the pipeline. I built such gates for a multi-agent system that serves a higher-education group with 700,000+ students, and the method is written up as a case study.

A larger organization asks a different question first: who may let an agent reach which system. An MCP gateway on Cloud Run is my published answer. It places the tools of a company behind one governed entry and leaves evidence after each call.

A company with data in many systems needs the layer underneath before any agent is useful. The site shows an analytics platform and a data warehouse fed from operational systems.

A company that sells to other companies by email can read how I treat deliverability, prospecting and funnels as engineering.

Several of the cities above are known for industries that the site does not cover. No case study on this site comes from energy, semiconductors, aerospace, banking, entertainment, biotech, ports or government. What transfers is the method under all of them: a process written down, an agent that reaches only the data its task needs, and tests that can stop a release.

What this is not

It is not a staffing agency and it is not a team. There is no bench, no account manager and no substitute: the person who answers your first message is the person who writes the code. If a project needs several engineers at once, I can be one of them, inside your team. I cannot be all of them.

How to start

Write to me with the role or the problem. For a role, the CV carries the history. For a contract, I ask to read the process and then reply in writing with a first piece of work and its limits. The pages for regions and cities say what a company in each of the cities above can expect.

Common questions

Are you in the United States?

No. I live in Monterrey, Mexico, and the work is remote. What a team in the United States gets is a shared working day and a senior engineer inside its own repositories, not a person in its city.

Is Monterrey on the same time as Texas?

From early November to mid March, yes: Austin, Dallas, Houston and San Antonio are on the same hour as Monterrey. From mid March to early November those cities are one hour ahead. Texas changes its clocks and Monterrey stays on UTC-6 all year.

How far is Monterrey from Eastern Time and from Pacific Time?

New York and Miami are one hour ahead of Monterrey from early November to mid March and two hours ahead for the rest of the year. San Francisco, Los Angeles and San Diego are two hours behind in the first period and one hour behind in the second. In both cases most of the working day overlaps.

Is this a nearshore staffing agency or an outsourcing firm?

Neither. I am one Senior AI Engineer and I work under my own name. There is no pool of engineers behind this page, so a company that needs many people at once needs something else.

Who keeps the code when a nearshore engagement ends?

The work is done in your repositories and your cloud accounts, so it is already with you. Ownership of the code is a term to put in writing before the work starts, together with the scope and the way access to your systems ends.

By city

Open to senior AI engineering roles and contract engagements, remote in the US, Mexico and Latin America.