Maksim Kobelev is a Senior Data Scientist based in Budapest with 7+ years in machine learning, specializing in NLP, computer vision and document processing, and five years of commercial experience delivering production-grade systems. He has led teams and driven architecture shifts from lightweight NN models to BERT-based and multimodal transformers, enabling few-shot on-premise learning and improved document clustering and extraction accuracy. At EPAM he built a GPU-accelerated, GStreamer-based OCR pipeline deployed to NVIDIA Orin edge devices for real-time package print-defect detection across global factories, and developed LLM-driven anomaly detection for large-scale general ledgers. Known for bridging research and operations, he combines model innovation (e.g., Visual-Transformer features, GCNs for receipts) with pragmatic deployment—Docker, Azure, Ansible—and operator-facing analytics. His background in financial technologies and applied mathematics helps him design robust, auditable ML solutions that reduce manual work and scale across enterprise environments.
5 years of coding experience
4 years of employment as a software developer
Master's degree, Financial Technologies and Data Analysis, 7.9, Master's degree, Financial Technologies and Data Analysis, 7.9 at Higher School of Economics
Contributions:2 releases, 40 pushes, 3 branches in 8 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.