What an engagement looks like

Custom software

You need a working web application built end-to-end, not a half-finished proof of concept. I deliver the whole thing: front-end, back-end, database, auth, deployment, documentation.

Best for: Founders, freelancers, and small teams who need a working system shipped, not an endless discovery process.

What you get

• React + Vite SPA with sensible routing, auth, and a polished UI
• Node.js / Express APIs with proper validation, rate limiting, and error handling
• MySQL or DynamoDB schema designed for the actual access patterns
• Repo handed over with deploy scripts, README, and a migration system
• One free week of bug-fix coverage after launch

Cloud & backend

Whether you need a fresh AWS deployment or a rescue on someone else's production fire, I handle the parts of the stack nobody else on your team wants to touch.

Best for: Businesses tired of paying twice the EC2 bill they should because nobody's audited it in two years.

What you get

• EC2 / nginx / PM2 / TLS setups with full deploy automation
• Serverless (Lambda / API Gateway / DynamoDB / S3 / CloudFront) where it fits
• Database migrations, backup strategy, and schema review
• Security hardening: rate limiters, security headers, audit logging, sane secrets handling
• Cost analysis with concrete numbers, not "it depends"

Game development

Unreal Engine work for indie projects and prototypes. Multiplayer systems, AI behaviour, and the web infrastructure to back a live game (player portals, telemetry, task queues).

Best for: Indie devs who want a non-gameplay engineer to handle the systems plumbing while they focus on the game.

What you get

• Unreal C++ and Blueprint development
• Steam multiplayer integration via Online Subsystem Steam
• Web portals that talk to running game sessions (Steam OpenID + auth tickets)
• Async play-while-away mechanics: players queue work for their characters via a web app
• Build/CI pipelines for game projects

Performance & reliability

Find where your stack breaks before your users do. Load testing, profiling, incident response, and post-mortem write-ups that actually get read.

Best for: Anyone whose support inbox is full of "the site is slow" tickets and they can't reproduce it.

What you get

• k6-driven load testing with live dashboards and percentile reporting
• CPU / memory / event-loop profiling on Node.js, Python, and game backends
• Identification of the actual bottleneck (usually not what the team thinks it is)
• Runtime toggles for safe production experimentation (rate limiters, maintenance mode)
• Incident response: dropping in during a fire and getting the system stable, then a written post-mortem