PlaleLab at the University of Oregon is an academic research lab carrying out research in computational and data infrastructure for AI, high performance and distributed computing, open science, provenance & reproducibility, and AI ethics and accountability.
Computational & Data Infrastructure for AI
We design and build cyberinfrastructure that makes AI/ML models easier to train, document, deploy, and govern.
High Performance & Distributed Computing
Our systems span the edge-cloud-HPC continuum, harmonizing resource-limited edge devices with centralized HPC and cloud resources.
Open Science
We build open, reusable tools and knowledge infrastructure for the broader research community.
Provenance & Reproducibility
We use provenance-based knowledge graphs to track how AI/ML models, datasets, and deployments relate to one another over time.
AI Ethics & Accountability
We research semi-automated ways to capture fairness, explainability, and behavioral information about deployed AI/ML models.