Postdoctoral Researcher — Agentic AI

1.0 FTE · CARC · MESA Goal 3 lead · UNM Jobs requisition 37416

The role

You will lead the agentic AI orchestration prototype for MESA, the NSF Integrated Data Systems & Services project based at CARC: self-hosted large language model serving on vLLM, retrieval-augmented generation (RAG) pipelines over curated scientific metadata, and the MESA Model Context Protocol (MCP) server framework — sandboxing, multi-agent orchestration, and evaluation against real science use cases from genomics to the Event Horizon Telescope. You will work directly with the principal investigator at CARC and with partner teams at the University of Arizona and UNC Chapel Hill / RENCI.

The prototype runs on national infrastructure that is already in place: CARC’s Easley (H100, L40S) and Hopper (A100) GPU clusters, the 5.2 PB iRODS-backed CyVerse Data Store, and ACCESS-CI and NAIRR allocations including the Jetstream-2 GPU cloud and the Open Storage Network. The MESA MCP stack is public and in use against the CyVerse Data Store today.

What you bring

A PhD in computer science or a computational science.
Systems-level Python.
An interest in how LLM agents should handle petabytes of scientific data responsibly.

What comes with it

A structured mentoring plan, first-author publications, and participation in the project’s TrustedCI security review.

Apply through UNM Jobs. The University of New Mexico is an Equal Opportunity Employer.