Backend & ML engineer. I build systems that turn messy real-world data into decisions — mostly in sports analytics, but the stack travels well.

Sports analytics — live in-game prediction engine for basketball: Bayesian Q-total models, TrueSkill 2 player ratings, WebSocket signal API with per-bot auth and a ready → sent → placed/skipped lifecycle.

Automation — bet placement automation across multiple bookmakers: iOS (browser automatisation + scripting), desktop (PC-side workers), Docker-isolated per-account environments.

Scraping — odds pipelines, live data ingestion, anti-detect patterns, browser simulation.

DevOps — live PostgreSQL migrations without downtime, Grafana + Prometheus + Loki observability stack, Caddy reverse proxy, Cloudflare Tunnel, Docker Compose on Hetzner VPS. Survived and remediated a crypto miner compromise (entry via exposed Docker port — don't do that).

Full-stack — Python, Go, PostgreSQL on the backend. React when the frontend can't be avoided.

Languages   Python · Go · SQL · Bash
ML          scikit-learn · CatBoost · Bayesian inference · Poisson-Gamma models
Backend     FastAPI · PostgreSQL · Redis · WebSocket
DevOps      Docker · Linux · Caddy · Grafana · Prometheus · Loki · Cloudflare
Scraping    Playwright · browser automation · anti-detect
Automation  iOS Shortcuts · desktop scripting · multi-account orchestration