Public sector procurement involves enormous datasets and complicated performance issues. The research team at University of California San Diego, led by Professor Vish Krishnan, wanted to leverage cutting-edge machine learning—including large language models (LLMs)—to find what drives contract overruns and management complications. With so much data to analyze and experiment with, their biggest need was scalable, reliable computing power over an extended project timeline.

By providing sustained access to our advanced Nvidia H100 GPU cloud computing resources, Canopy Wave made it possible for the research group to:

• Quickly ramp up novel machine learning models, including their custom predictive analytics pipeline and causal inference techniques.
• Experiment with LLMs to deepen insights, even as research needs shifted during academic development.
• Manage project workflows reliably through relocation, new team members, and the inevitable pivots that come with rigorous research.

Our flexible, secure, and high-performance AI cloud computing gave the University of California San Diego team freedom to explore ideas that might otherwise be slowed by technical barriers. They could ingest high-dimensional contract datasets, run robust double machine learning models, and iterate rapidly to refine their predictors and metrics for contract performance and overruns.

Although the core research findings are still under wraps until publication, Professor Krishnan and his team have highlighted how Canopy Wave's support was “critical” in sustaining project momentum and enabling deep methodological innovation. Our H100 GPU cloud computing resources allowed them to meet academic milestones and continue pushing boundaries in contract analytics—setting the stage for future policy recommendations that could shape public sector procurement for years to come.

This collaboration not only drives academic progress but also demonstrates how Canopy Wave's customer commitment can energize world-class research, create real flexibility for evolving projects, and help solve problems with national impact.