Time and Venue
3 October 2025, 1:30 to 3 PM (ET)
Klaus 1116, Klaus Advanced Computing Building (KACB), Georgia Tech
Title and Abstract
Inference & Agentic Systems for Science
Large Language Models (LLMs) are increasingly being adopted as powerful tools in scientific research to accelerate workflows. In this tutorial, we will explore how graduate students in quantum chemistry (and related disciplines) could leverage LLMs not only as coding assistants but also as agentic tools for domain-specific tasks. This session will emphasize practical skills: participants will use gemini-cli and/or opencode to interact with agentic AI directly from the terminal. Through guided exercises, students will learn how to engage in better prompting, employ slash commands, and develop/connect MCP (Model Context Protocol) servers. Examples will highlight workflows such as analyzing quantum chemistry datasets, reproducing figures from the literature, and developing MCP tools tailored to domain-specific tasks in quantum chemistry. By the end of the session, participants will have hands-on experience with modern AI tooling and learn to leverage freely available technologies as students at Georgia Tech.
Instructors
Ph.D. Student,
School of Chemistry & Biochemistry,
Georgia Tech
Austin is a PhD candidate in Chemistry at Georgia Tech in the Sherrill group, studying non-covalent interactions using quantum chemistry methods. As a National Science Foundation Graduate Student Fellow, he has worked on benchmarking and lowering the cost of symmetry-adapted perturbation theory (SAPT) methods to be used in downstream machine-learning applications on larger molecular systems. During this time, he has furthered his passion for high-throughput quantum chemistry through developing MPI-distributed software for job management and incorporating QCFractal/QCArchive for long-term data standardization, computation, and storage. As a MolSSI Software Fellow, he is working on implementing an accurate functional group-based SAPT in Psi4, an open-source quantum chemistry package, to more effectively train atomic-pairwise neural networks. In collaboration with the ARTESAN group at Georgia Tech, he has been working on developing an AI4QC framework to build a computational chemistry assistant to assistant both experts and non-experts alike to correctly select an appropriate level of theory and run quantum chemistry calculations to answer research questions through computation.
Ph.D. Student,
School of Computer Science,
Georgia Tech
Piyush Jha is a Ph.D. student in Computer Science at Georgia Tech advised by Prof. Vijay Ganesh. His research focuses on AI for Science and neurosymbolic techniques, developing computational methods that combine machine learning, symbolic reasoning, and reinforcement learning to discover new physics theories and design new materials with targeted properties. Piyush brings industry and research experience from Amazon Science, where he designed LLM-based autonomous model-building agents, and from Quadrical AI, a leading solar AI startup where he developed predictive analytics and digital-twin solutions for renewable energy systems. He has also contributed to multiple academic labs, building high-performance tools at the intersection of AI and scientific discovery. A two-time SMT-COMP winner, he has received the University of Waterloo Graduate Scholarship and the Pasupalak Scholarship in Robotics and AI, and his work has been published in top venues including ACM Transactions, IJCAI, AAAI, and ECAI.
