At D2L, I’m a core owner of the generative AI suite for Brightspace, shaping feature design, output quality, observability, and model integration. I designed and delivered instructor and student-facing LLM tools, balancing quality, latency, cost, privacy, permissions, and regional constraints.
Working on brightspace’s first LLM-backed AWS serverless services, integrating Node.js capabilities with C# workflows and establishing end-to-end LLM observability. Earlier, I modernized legacy C# MVC functionality into Lit web components and researched collaborative-filtering recommendation models in PyTorch at the University of British Columbia.
