Middle Pythoni arendaja

- Participated in the design of high-load REST API modules using FastAPI and asyncio, which increased service throughput by 35% and reduced response time from 400ms to 120ms. - Integrated ML models (NumPy, Pandas, PyTorch) into the production environment for the bank's scoring system. Optimized the data processing pipeline, reducing the time to generate business reports by a factor of 4 (from 40 minutes to 10 minutes). Optimized SQL queries using SQLAlchemy in PostgreSQL for a document recognition system at a logistics company, configured indexing and caching in Redis, which reduced the database CPU load by 25%. Configured automated deployment pipelines (Docker Compose, GitLab CI/CD), reducing Time-to-Market by 15% and eliminating manual deployment errors.