Summary
Fraser Greenlee is a Principal Software Engineer based in Scotland with a decade of experience building production ML systems and data platforms across startups and enterprise teams. He moved from applied ML roles at Cohere—where he led reward-model research, automated large-scale human evaluations, and co-authored work on synthetic critiques—to senior engineering and platform roles at Convergence and now Salesforce. Fraser has a strong track record of turning prototypes into production (from Pyspark DAGs and tokenizers to JAX training infrastructure and automated eval tooling) and has repeatedly driven SOTA improvements in preference modelling. He combines hands-on implementation (Python, JAX, Pyspark, tokenizer work) with cross-team leadership, regularly presenting evaluation results to boards and standardizing tooling used company-wide. Comfortable in both research and production contexts, he has deep experience scaling ML workflows, datasets, and evaluation pipelines. A practical problem-solver, he often bridges gaps between annotation, engineering and research to make models both competitive and product-ready.
10 years of coding experience
6 years of employment as a software developer
Bachelor's degree Computer Software Engineering, Bachelor's degree Computer Software Engineering at University of Glasgow
English