Backend engineer focused on clean, reliable APIs and backend systems that hold up under real usage — not just demos.

I’ve been building backend software since 2020, starting with pragmatic CRUD-based APIs and growing into maintaining and refactoring long-running systems. My focus is on correctness, clear system boundaries, and architectures that can evolve without breaking under change.

I enjoy working close to the metal when needed, using Rust and C for performance-critical paths, while leveraging higher-level tools where they make sense.

stexs-spike is a backend platform for multiplayer games, designed to provide authentication, user management, and inventory systems as long-running, production-grade services.

The current iteration includes:

• Authentication and user accounts
• Player inventories and cross-inventory systems
• OAuth2 integrations
• A web platform for profile management, social features, item purchases, and achievements

Many of these features are already implemented. Ongoing development continues alongside a next-generation architecture that incorporates lessons learned from maintaining and refactoring real backend systems over time.

STEXS is a backend infrastructure platform for multiplayer games, designed to provide authentication, inventory, storage, and operational primitives as shared services — without vendor lock-in or fragmented integrations.

The platform is built backend-first, with architecture defined before feature scale.

Key highlights:

• Rust-based core with strong type safety and async workflows
• Intent-driven APIs (not CRUD), designed around real business actions
• Vertical Slice Architecture with clear domain ownership (Auth, Inventory, Storage, etc.)
• Event-driven core using Redpanda for loose coupling and auditability
• Distributed SQL via YugabyteDB with ACID guarantees
• Garnet for low-latency caching and atomic operations, chosen for significantly higher throughput than Redis on the same hardware (even with multithreading enabled)
• Pingora at the edge for request validation, rate limiting, and JWT caching

The goal of this iteration is correctness, clarity, and long-term evolvability — building infrastructure that can grow without breaking existing integrations.