Research Engineer · LLM Evaluation · Montreal, QC
I build scalable evaluation infrastructure and benchmarking systems to understand how language models fail, where they generalize, and what it takes to make them more reliable. Currently finishing my MSc at McGill University, where my thesis focuses on evaluation methodology and the limits of standard benchmarks for LLMs. Advised by David Adelani. Previously Research Engineer at Lelapa AI, and Data Scientist at Okra Technologies.
Seeking roles building evaluation systems that improve how language models are understood, measured, and trusted.
Built and drove adoption of end-to-end evaluation infrastructure — covering benchmark development, job orchestration, failure-mode diagnostics, and continuous regression monitoring across 10+ speech and language models. Worked cross-functionally as an internal evals-as-a-service layer supporting research, engineering, and product teams; reducing iteration cycles by 40%.
Built and shipped production ML systems for fraud detection, NLP classification, and financial behaviour modelling. Designed end-to-end ML pipelines with automated retraining, experiment tracking, and API deployment.