Profile
I build software systems that are meant to survive contact with reality.
I work with companies on platform architecture, AI integrations, and data systems. Most of the work sits in the messy middle between a promising idea and a production system that a team can actually run, maintain, and improve over time.
How I help
The projects vary, but the pattern is usually the same: make the system simpler, more reliable, and easier for the team to own after the initial build is done.
AI & LLM systems
I help teams design agents, RAG systems, and LLM features that hold up outside a demo. That means working with messy data, real user behaviour, and the production constraints that usually show up later than they should.
Platform architecture
I design backends and infrastructure with enough structure to last. Clear boundaries, sensible interfaces, and the operational visibility a team needs to keep shipping without slowly making the system harder to work on.
Data platforms
I build pipelines, warehouses, and analytics foundations that people can actually rely on. The goal is straightforward: give the business useful data, and give engineering something they can keep extending as the company grows.
Operational leverage
I also spend a lot of time on automation and internal tools. Small systems that remove repetitive work often end up being some of the highest-return projects for a lean team.
The stack I reach for
I am not overly attached to tooling, but these are the technologies I use most often. Python and TypeScript for application work, Postgres, ClickHouse, Elasticsearch, and OpenSearch for data, Airflow and Kafka for orchestration and streaming, and the usual set of modern model providers when AI is part of the job.
- Python
- TypeScript
- Go
- FastAPI
- Node.js
- React
- Next.js
- Tailwind
- GraphQL
- Postgres
- MongoDB
- ClickHouse
- Elasticsearch
- OpenSearch
- Redis
- Airflow
- Kafka
- Docker
- Kubernetes
- AWS
- Linux
- Grafana
- OpenAI
- Anthropic
What drives me
I like the kinds of problems that sit across engineering, product, and operations. The interesting part is usually not just writing code. It is figuring out where better systems, better automation, or better data flow will actually change how a team works.
These days that mostly means system design, architecture planning, developer workflow improvements, and AI work that needs to be useful in production rather than impressive in a prototype.
If this sounds close to what you're building
Most useful conversations start with a system that has become harder to scale, an architecture decision that needs a second opinion, or a prototype that now has to become a real product. If you are in that kind of spot, GitHub is the easiest place to reach me.