Remote AI engineer for companies in Raleigh and Durham, NC

I am a Senior AI Engineer and I work remotely from Monterrey, Mexico, with companies in Raleigh and Durham, as a contractor or in a senior role. I am not in the Research Triangle, and no case study on this site comes from biotechnology or clinical research, so this page says exactly what transfers and what does not.

Raleigh and Durham form the Research Triangle, known for biotechnology and pharmaceuticals, for clinical research, for its universities, and for software and networking. Much of that work happens in laboratories and in studies, where a wrong answer is expensive and an invented one is worse. This page describes, sector by sector, what an AI engineer in Mexico can do for a company of that kind and where he stops.

Laboratories and the documents they run on

A laboratory keeps what it knows in protocols, methods, notebooks and reports. People in biotechnology and pharmaceutical companies tend to want one thing from AI before any other: to put a question to that material and receive an answer that names its source, or an honest "not found". That work is described under RAG and data platforms.

No case study on this site comes from biotechnology or pharmaceuticals. What transfers is a system of the same shape in another field. For a lean manufacturing practice I built an open corpus of documentation and an assistant that answers from it alone, and the case study shows how the search behind it combines meaning and exact words. A laboratory's corpus is private and controlled. That changes who may read what, and it does not change the method.

Clinical research

Organizations that run studies answer for every step: who changed a record, when, and on what basis. An agent in that setting has to leave the same kind of trail, and a change to it has to pass tests before it reaches anyone. No case study on this site comes from clinical research. One does come from a laboratory of another kind: a COVID-19 testing platform, with a role for each operator and a confirmation at every stage from registration to result. It is operations software that I built in 2021 and 2022, and it is not an AI system. What transfers is the guide on agent auditability, which sets out the evidence an agent should leave, and the release gates described further down. I do not interpret the rules on validated systems or on patient data. Your quality team writes the requirement, and I build to it.

Software, networking and the universities

Software and networking companies usually have an agent in a product already. Their question is whether tomorrow's version can ship. That is LLM evaluation, the center of my published work: a short set of cases on every merge, the full regression before a version is promoted, and each case run as its own synthetic user so that one test cannot leak into the next. My open-source MCP servers and my research on agents are public for the same readers.

Universities are the one sector here with a close match. Those gates belong to a multi-agent system that serves a higher-education group with 700,000+ students. It answers questions about admissions and enrollment, schedules appointments and sends each person to the right human team.

Eastern time against Monterrey's

Monterrey does not change its clock. From early November to mid March, Raleigh and Durham are one hour ahead of Monterrey. From mid March to early November they are two hours ahead, on the same Eastern hour as New York. I work inside your repositories and your cloud accounts, and review happens on the code and in writing, with a call when a decision calls for one.

Before you write

Tell me the role or the problem. I will read how the work is done now and reply with a first piece to build, the access it requires and the things I would not promise. I take senior roles and contract engagements. I am not a person in the Triangle and not an authority on your regulation. If the work needs either one, this is the wrong page.

Common questions

Do you live in the Research Triangle?

No. I live and work in Monterrey, Mexico. Companies in Raleigh and Durham would work with me remotely, through shared repositories, written reviews and calls.

When it is noon in Durham, what time is it in Monterrey?

Eleven in the morning from early November to mid March, and ten in the morning from mid March to early November. Raleigh and Durham are one hour ahead of Monterrey in the first period and two hours ahead in the second, because Monterrey does not change its clock.

Is there published work from biotechnology or clinical research?

No case study on this site comes from biotechnology, pharmaceuticals or clinical research. One describes a COVID-19 testing platform for a laboratory, which is operations software built in 2021 and 2022, not an AI system. What transfers is the method: answers drawn only from a controlled set of documents, a record of what an agent did, and tests that decide whether a change is released.

Can you build for a validated system or for study data?

I build to the requirements that your quality and privacy teams put in writing, and I do not interpret those rules for you. Development and testing can run on scripted cases with mocked state and synthetic users, which is how the evaluation work in my case study is built.

Is anything on the site close to a university?

Yes. The case study on evaluation gates describes a multi-agent system that serves a higher-education group: questions about admissions, enrollment, scheduling and routing to the right human team. It is the published work closest to a university's own operations.

By city

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