Machine Learning Research Engineer
Building efficient, production-ready AI systems that bridge the gap between novel research and real-world impact.

My work revolves around the full lifecycle of machine learning models, from architectural design to deployment and optimization. I'm particularly interested in Efficient Model Architectures:

I am currently working on increasing the efficiency of Hypencoder models (retrieval models) by reducing the number of active parameters using multi-objective optimization techniques. So far, I have achieved 7x reduction in the number of active parameters and 6x increase in document scoring throughput, with performance comparable to SOTA retrieval models.

Reach out if you would like to discuss anything related to my interests.