Postdoctoral Fellow: Agentic AI for Scientific Discovery

Location: University of Waterloo, David R. Cheriton School of Computer Science or Vector Institute, Toronto

We are seeking a postdoctoral fellow to join a large-scale project in collaboration with industry partners on AI for scientific discovery. The project's goal is to develop a deep research assistant to accelerate scientific R&D workflows in natural science, starting with an initial exploration in Chemistry. This involves solving fundamental challenges in how AI systems retrieve complex information, reason over it, and autonomously interact with scientific data and software. You will join a dedicated team of PhD and MSc students, co-mentoring them while leading one of the project's core research thrusts.

Research Areas

The postdoctoral fellow will be expected to lead the project's primary research thrusts. We are looking for candidates with expertise in one or more of the following research areas:

Multimodal Information Retrieval: Developing novel retrieval frameworks that unify heterogeneous scientific data (text, tables, images, time series) from data lakes.
Compositional Reasoning: Investigating new training methodologies to teach models to perform complex, multi-hop reasoning with provenance.
Robust Agentic AI: Building generalist, multimodal agents (VLMs) that can quickly learn to operate proprietary scientific software. This thrust has a major focus on agent safety and robustness.
Data-Centric AI for Agents: Creating scalable, data-centric frameworks for agent training and validation. This includes research into synthetic data, large-scale automated evaluation using simulated users and human-preference Arenas.

Role and Responsibilities

Lead a core research thrust in one of the areas listed above.
Publish novel, high-impact research at top-tier AI conferences (e.g., NeurIPS, ICML, ICLR, ACL).
Collaborate closely with the PIs and the multidisciplinary team of AI researchers and domain (chemistry) experts.
Co-mentor a team of dedicated PhD and MSc students.
Contribute to the development and open-sourcing of integrated prototypes and associated benchmarks.

Preferred Qualifications

A completed (or near-completed) PhD in Computer Science, Chemistry, or a related field.
A strong publication record in one or more of the following areas: Natural Language Processing (NLP), Machine Learning (ML), Multimodality, Chemistry.
Strong programming skills (Python, PyTorch/JAX).
Experience with agentic AI, tool-use, or reinforcement learning.
Experience with multimodal models (VLMs) or GUI automation.
Experience with data-centric AI, evaluation, or synthetic data.

What We Offer

A competitive salary and benefits package.
Dedicated computational resources (dedicated GPU cluster).
Direct, deep collaboration with both academic PIs and a highly engaged industrial partners.
Mentorship opportunities and the chance to lead a small team of graduate students.
A clear path to high-impact publications and a real-world, deployed system.

How to Apply

Please send the following to Victor Zhong ([email protected]) with the subject line "Postdoc Application: Agentic AI for Science":
A Curriculum Vitae (CV).
A brief cover letter (1 page max) outlining your research interests and specifying which of the four research areas you are most interested in.
Names and contact information for 2-3 references.

The University of Waterloo is committed to fostering an equitable and inclusive environment. We encourage applications from all qualified individuals, including women, Indigenous peoples, persons with disabilities, and members of visible minorities.