Research Aide– DSL – Ishtiaque Mahmud, Quazi – 3.5.26

Argonne National Laboratory•Lemont, IL

4d•$31 - $47

About The Position

This project focuses on implementing and scaling a multi-agent system (MAS) training and evaluation pipeline on Aurora, as part of the AuroraGPT initiative at Argonne. I am a postdoctoral researcher partially funded by ALCF, working with Dr. Venkatram Vishwanath and Dr. Rajeev Thakur on large-scale LLM systems and evaluation. The goal is to translate a recent multi-agent research framework into a practical, runnable system on Aurora that can support large-scale experimentation and future scientific workflows. The MAS framework studies how multiple LLM agents collaborate to solve complex tasks, and how system-level evaluation signals can be transformed into agent-level and message-level training signals for improving cooperation, reliability, and efficiency. The summer student will assist with implementing the multi-agent orchestration, logging and trace collection, evaluation hooks, and scalable execution on Aurora, enabling controlled experiments and benchmarking at leadership scale. This work aims to demonstrate one of the first end-to-end multi-agent LLM systems running natively on Aurora and contribute toward publishable results in multi-agent learning for science.

Requirements

The entirety of the appointment must be conducted within the United States.
Applicants must be: Currently enrolled in undergraduate or graduate studies at an accredited institution.
Graduated from an accredited institution within the past 3 months; or
Actively enrolled in a graduate program at an accredited institution.
Must be 18 years or older at the time the appointment begins.
Must possess a cumulative GPA of 3.0 on a 4.0 scale.
Must be a U.S. citizen or Legal Permanent Resident at the time of application.
If accepting an offer, candidates may be required to complete pre-employment drug testing based on appointment length.
All students remain subject to applicable drug testing policies