Summary
Mikhail Nefedov is a data scientist with a decade of experience and a strong NLP foundation, currently applying his skills at Miro in Amsterdam. He blends industry work—building production-ready models for call-centers, voice assistants, fraud detection, and CRM scoring—with four years of university-level teaching, having designed and taught a highly rated NLP course. His projects balance practical engineering (vector search, real-time call-topic classification) with research-driven techniques like knowledge distillation for compact text embeddings. Comfortable across the ML stack, he’s solved problems ranging from grammar-based entity extraction to gradient-boosting and deep learning models, and he prioritizes interpretability and robustness when data is scarce. Relocating to the Netherlands, he brings a linguistics background that informs clever, language-aware solutions beyond standard ML approaches.
10 years of coding experience
6 years of employment as a software developer
Master's degree, Natural Language Processing, Master's degree, Natural Language Processing at Higher School of Economics
English, Russian