About Me
Building systems that
think, scale, and endure.
I am a software engineer specialising in agentic AI systems, distributed cloud infrastructure, and immersive full-stack web. I obsess over the intersection where autonomous intelligence meets production reliability.
By the numbers
Toolbox
The primary technologies I reach for across the three pillars of my practice.
Agentic AI
Microsoft AutoGen, LangGraph, LangChain, OpenAI API, Ollama, OpenLit, DeepEval, RAG pipelines
Cloud & Infrastructure
Kubernetes, Docker, Kafka, MQTT, TimescaleDB, PgBouncer, MinIO, Nginx, Cloudflare
Full-Stack & Spatial
Next.js 15, Astro, TypeScript, PostgreSQL, Prisma, Three.js / WebGL, Unity, Unreal Engine 5
How I work
The principles that guide every project from first commit to production.
- 1
Systems thinking first
Before writing a line of code I map data flow, failure modes, and scaling boundaries. Good architecture is invisible when it works and obvious when it doesn't.
- 2
Observable by default
Every system I ship is instrumented from day one - structured logs, distributed traces, and cost attribution so problems surface before users notice them.
- 3
Ship, then iterate
A working system in production beats a perfect system in planning. I favour tight feedback loops, feature flags, and zero-downtime deploys over big-bang releases.
Selected highlights
Multi-agent ecosystem with full LLM observability
Dynamic skill-loading architecture across an autonomous agent swarm, instrumented end-to-end with OpenLit for latency, token, and cost telemetry.
Full-cycle SaaS ERP delivery
Sole engineer - architecture through production - for a multi-tenant music school platform using Next.js 15, PgBouncer, and MinIO.
Enterprise Kafka + MQTT SCADA pipeline on Kubernetes
Replaced a monolithic relay with an event-driven microservices architecture capable of ingesting thousands of sensor events per second.