AI Research Post Doctoral Fellow

The University of New Mexico's Center for Advanced Research Computing (CARC), within the Department of Computer Science, seeks a full-time Postdoctoral Researcher to lead development of an open-source agentic artificial intelligence platform as part of a federally funded, multi-institution research initiative. The Postdoctoral Researcher will design and build the project's agentic AI stack-open-weight large language models served at scale, retrieval-augmented generation (RAG) pipelines, a Model Context Protocol (MCP) server framework, sandboxed execution, and multi-agent orchestration-and will direct a distributed engineering effort spanning the collaborating institutions. The position is supervised by and co-located with the Principal Investigator at CARC, with secondary mentorship from collaborating co-investigators at partner institutions. All work follows open-source, reproducible-research practice.

Primary Duties and Responsibilities

1. Leads the design, development, and evaluation of the project's agentic AI platform, including the serving of open-weight large language models (e.g., vLLM-served models), retrieval-augmented generation pipelines, the Model Context Protocol (MCP) server framework, sandboxed code execution, and multi-agent orchestration.
2. Directs and coordinates a distributed engineering effort, leading regular technical meetings with the partner-institution team and graduate research assistants, and presenting at design reviews and project milestones.
3. Conducts benchmarking and performance evaluation of LLM serving and agentic workflows on high-performance GPU systems (e.g., H100 / A100 / L40S) and national cloud allocations, and documents empirical hardware and performance findings.
4. Leads and contributes to peer-reviewed, open-access publications (target of at least two first-author papers), and disseminates results through public code repositories, containerized reproducible workflows with persistent identifiers (DOIs), and FAIR data practices.
5. Participates in security and responsible-AI review activities, including prototype security review and engagement with the project's external AI ethics advisory board.
6. Co-teaches research-computing and data-science training workshops (e.g., R, Python, Linux, ML/AI pipelines) and contributes training modules to the project's education and workforce-development activities.
7. Co-mentors graduate research assistants contributing to the agentic AI and MCP workstreams.
8. Participates in the annual program meeting and represents the project's technical progress to collaborators, sponsor program staff, and the broader research community.
9. Contributes to grant reporting and to the preparation of follow-on proposals, including empirical hardware-specification and benchmarking content.
10. Performs related duties as assigned in support of the project's goals and the Fellow's professional development.

Mentoring and Professional Development

Consistent with UNM's expectations for postdoctoral training, the Fellow and mentor will jointly prepare an Individual Development Plan (IDP) within 30 days of hire, organized around the National Postdoctoral Association core competencies, with semiannual review. The Fellow will receive weekly one-on-one mentorship from the PI, structured career advising across academic, national-laboratory, and industry pathways, grant-writing experience, and visibility through the project's national partner network. The Fellow will complete UNM's Responsible Conduct of Research (RCR) training within the first six months.

Due to budgetary constraints, we are unable to sponsor or take over sponsorship of an employment Visa. Applicants must be authorized to work in the United States on a full-time basis.

Minimum Qualifications:

• Ph.D. (or terminal degree) in Computer Science, Artificial Intelligence, Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, Computer Engineering, Computational Science, Data Science, Management Information Systems, Information Science, or a closely related field, completed by the date of appointment.
• Demonstrated research experience in machine learning, applied artificial intelligence, distributed systems, or research software engineering, as evidenced by publications, software, or other scholarly products.
• Programming proficiency in one or more relevant languages (e.g., Python, Rust, Go, C/C++, JavaScript/TypeScript, or R)

Preferred Qualifications:

• Experience with large language models, including model serving (e.g., vLLM), retrieval-augmented generation, agentic/multi-agent frameworks, or the Model Context Protocol (MCP).
• Experience developing and deploying containerized, reproducible workflows (e.g., Docker, Kubernetes/Helm) on HPC or cloud infrastructure (e.g., SLURM, OpenStack, ACCESS-CI resources).
• Experience building APIs and services (e.g., FastAPI, OpenAI-compatible inference endpoints) and integrating authentication and orchestration tooling.
  Track record of open-source software development, code review, and FAIR/open-science practice (public repositories, DOIs, reproducible pipelines).
• Experience leading or coordinating distributed teams, mentoring students, or teaching technical workshops.
• A demonstrated commitment to cultivate an understanding of the rich and varied cultures of New Mexico and to the success of the university's mission to serve local and global communities

Only applications submitted through
the official UNMJobs site will be accepted. If you are viewing this job
advertisement on a 3rd party site, please
visit UNMJobs to submit an application.

Applicants who are appointed to a UNM faculty position are required to provide an official certification of successful completion of all degree requirements prior to their initial employment with UNM.

The University of New Mexico is committed to hiring and retaining a diverse workforce. We are an Equal Opportunity Employer, making decisions without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, veteran status, disability, or any other protected class.