JAMIE BARTON
Atleast 20 years of building the systems that make code ship.
I design and build coding systems: the compilers, pipelines, platforms and tools that other engineers rely on every day. Two decades in, still writing the hard parts myself.
See the systems I've builtTHE JOURNEY SO FAR
Two decades, written one system at a time.
I started coding on hardware that measured memory in megabytes, shipping small utilities in C and then Java when the JVM was still the exciting new thing. Those early years taught me the discipline that still shapes how I work: understand the machine, respect the constraints, and never ship a system you cannot reason about. Through the 2000s I moved deeper into the plumbing. I built parsers, task schedulers and the internal tooling teams actually depend on. When the web scaled up, I scaled with it: distributed queues, service meshes before they had that name, and the observability needed to keep them honest under load. The last stretch has been about leverage. I build systems that let other engineers move faster without breaking things: type-safe build pipelines, code generators, and platforms that turn a week of setup into a single command. Twenty years in, the tools change but the job is the same, make the complex reliable.
SYSTEMS I'VE BUILT
Real infrastructure, still running.
Build pipeline engine
A distributed, cache-aware build system that turned 40-minute CI runs into two. Content-addressed artifacts, remote execution, and deterministic outputs across every machine.
Streaming ingest platform
Handled millions of events per minute with backpressure that never dropped a message. Exactly-once semantics, replayable topics, and dashboards engineers actually trusted at 3am.
Type-safe code generator
One schema, many targets. It produced clients, servers and docs in sync so an API change could never silently break a consumer. Saved the team from a whole class of runtime bugs.
Developer platform
A self-service platform that provisioned environments, secrets and databases from a single command. New services went from a week of setup to production in an afternoon.
Query planner and cache
A cost-based planner sitting in front of a heavy datastore. It rewrote expensive queries and cached the right slices, cutting p99 latency by an order of magnitude.
Observability toolkit
Tracing, metrics and structured logs wired through every service by default. When something broke, the answer was already on a graph instead of buried in a grep.
SKILLS AND STACK
What I reach for.
Languages
- Go and Rust for systems work
- TypeScript across the stack
- Java and the JVM ecosystem
- C for the parts close to the metal
- Python for tooling and glue
Infrastructure
- Kubernetes and containers
- Kafka and event streaming
- PostgreSQL and Redis
- Terraform and infra as code
- AWS and GCP
Practice
- Distributed systems design
- Compilers and code generation
- CI/CD and build tooling
- Observability and profiling
- Mentoring and code review
ABOUT JAMIE
A builder, not a bystander.
I'm Jamie Barton, a systems engineer who has spent 20 years atleast writing the code underneath the code. I care about correctness, performance and the kind of clean abstractions that survive contact with a real team. I still read the source, still profile before I optimise, and still believe the best system is one other people can extend without asking me. If you're building something hard, I'm the engineer who enjoys the hard part.
COMMON QUESTIONS
Working with me.
What kind of work do you take on?
Deep systems work: build pipelines, data platforms, code generators, query planners, observability and the internal tooling teams depend on. If it sits underneath the product and needs to be reliable, that is my patch.
Do you write code yourself or just advise?
I write the hard parts myself. Twenty years in, I still read the source, profile before I optimise and ship the tricky pieces rather than hand them off.
Which languages and stacks do you use?
Go and Rust for systems work, TypeScript across the stack, Java and the JVM, C close to the metal and Python for glue. On infrastructure I lean on Kubernetes, Kafka, PostgreSQL, Redis, Terraform and both AWS and GCP.
Do you work remotely or on site?
Both. Most of the work is remote, and I am happy to spend time on site for kickoffs, architecture reviews or when a team needs someone in the room.
Can you help my engineers, not just deliver code?
Yes. A lot of the value is leaving a team able to extend the system without me. I mentor, review code and document decisions so the work outlives the engagement.
How do I start a conversation?
Tell me the hard problem you are stuck on. Email me a short description of the system and the constraints, and I will come back with an honest read on whether I am the right fit.
Jamie Barton. 20 years atleast of building coding systems.