Summary
Vasilii Karmazin is a Software Engineer III based in Prague with a decade of experience building production-grade AI and backend systems, currently contributing to Stats Perform. A Python-focused engineer, he has deep hands-on experience in NLP dialogue systems, similarity search, and deploying GPU-backed services, having supported large-scale chat services for millions of customers at Tinkoff. He designs and maintains microservice architectures with FastAPI, SQLAlchemy, Elasticsearch, Kubernetes and CI/CD pipelines, and partners closely with ML teams to move models from research into reliable production. His background blends ML research (Master’s in Machine Learning and Data Processing) with practical ML engineering—implementing PyTorch-based intent recognition, A/B testing, and observability for critical customer-facing features. Curious about broader sensory AI, he’s actively exploring computer vision, audio processing and 3D applications, bringing both production discipline and research-minded experimentation to projects.
10 years of coding experience
5 years of employment as a software developer
Master's degree, Machine Learning and Data Processing, Master's degree, Machine Learning and Data Processing at Masaryk University Brno
Bachelor's degree, Software Engineering, Bachelor's degree, Software Engineering at National Research University of Electronic Technology (MIET)
English, Russian, Czech