Summary
Rayan Nathoo is a Machine Learning Engineer with a decade of experience building research-driven and production-ready ML systems, currently focusing on Bayesian optimization and low-data regimes at Atinary in Lausanne. He combines strong academic grounding from EPFL with hands-on expertise across audio ML, computer vision, NLP, model compression and DSP, having shipped prototypes and production pipelines for companies like Bose, Logitech and European broadcast projects. Rayan has published work on tiny speech enhancement via knowledge distillation (ICASSP 2024) and helped deploy real-time models in studio environments, reflecting a knack for translating novel research into robust, low-latency solutions. He’s equally fluent in infrastructure and developer experience—designing scalable training and inference platforms using MLflow, AWS Batch, Docker and MLServer. Colleagues describe him as a collaborative problem-solver who thrives on constrained, impactful applications spanning audio, vision, language and biology. Outside typical ML roles, he blends product-minded engineering with research curiosity, often bridging prototyping and long-term maintainability.
10 years of coding experience
3 years of employment as a software developer
Scientific baccalauréat, Major in Mathematics, Scientific baccalauréat, Major in Mathematics at Lycée français Jean-Mermoz, Dakar
Master, Computer and Communication Sciences, Master, Computer and Communication Sciences at EPFL
French, English, Spanish