At EPAM Systems, I develop and validate machine-learning models for client projects, from data preparation through evaluation. I build reproducible training and inference pipelines and implement evaluation systems to compare model versions.
I also optimize training and inference performance and debug numerical and system-level failures. In my daily workflow, I use Claude Code for implementation, debugging, refactoring, and test writing; my Anthropic certification covers LLM agents and MCP tool integrations.

