I designed, built and now run a multi-tenant event-bus platform from zero — AI-assisted, end to end. ~10 years in backend; I moved to Go as my primary language 3 years ago. I can take a system from nothing to production single-handedly, and hold the line on quality with TDD, contract tests and a code-review gate.
Nearly 10 years in backend, across microservices, API platforms, real-time messaging and reliable delivery. I've built core backends for a large sports/content platform (members, payments, auth, content, API gateway — 5+ subsystems) and shipped real third-party payment integrations and wallet settlement. The one thing that defines me: I can deliver a system end to end on my own — from environment setup and development to launch and production ops. Exactly what a remote, high-autonomy team needs.
Designed and built from zero, AI-assisted: it keeps Kafka behind a gateway and exposes only HTTP / SSE / WebSocket / MCP, so apps and AI agents can "subscribe to an event stream in one line." Auth, rate-limiting and multi-tenant isolation are all handled by the platform. Built as a Go monorepo modular monolith — a single binary to cut ops cost, while keeping clear service boundaries and a ready Compose/Helm split path.
Top-up/withdrawal integration with real third-party payment gateways (incl. card verification); wallet uses transactional locking to prevent over-withdrawal and double credits; agent commission settlement.
Extended and customized OpenIM: rearchitected push from broadcasting to all gateways O(N) to a Redis connection index + targeted delivery O(1), fixing traffic amplification under k8s scale-out.
Built a Go plugin for APISIX (multi-tenant IP allow/deny, maintenance windows); several third-party logins (Apple multi-audience, PKCE); AI content generation integration.
Self-run homelab: Proxmox LXC dev environments + shared middleware; VPS deploys via Docker Compose, reverse-proxy automation, TLS, fail2ban/UFW hardening.