Working with me

I’m an IT consultant. I help systems and teams work together to solve real-world problems. I’ve done this as an engineer, team lead, and CTO, across payment processing, distributed systems, and data platforms.

You can hire me to scope a problem, develop a first solution, or take responsibility for a larger piece of work. I work across product requirements, architecture, implementation, and the way teams organize around a system.

How I approach architecture

1. Scope the problem. Understand what needs to work, who depends on it, and the constraints: product requirements, existing systems, people, time, and cost.
2. Draft a first solution or MVP. Build enough to test the assumptions and learn what the problem actually requires.
3. Decide what needs to last. Identify the long-lived components and the short-lived parts based on the problem and product lifecycle. A component that supports a durable business need and an experiment that may disappear next month should be designed with different lifetimes in mind.

I use clear interfaces and isolation so those parts can evolve or be replaced independently. These boundaries need revisiting as the product changes. They also need owners: the architecture has to work for the people building and running it.

This builds on the notes I wrote about design decisions in 2019.

Typical engagements

Platform Architecture

A platform needs to change, but other teams depend on it. I help work out the boundaries, interfaces, and migration steps, then build the changes with your engineers.

My starting point is that we will get some decisions wrong. There should be a way to replace a component without having to replace everything around it. I wrote a few notes about this in 2019.

Infrastructure & Cloud

Deployment, scaling, or day-to-day operations are taking more effort than they should. I work on the infrastructure and automation, including production debugging and migrations.

Most recently that has meant AWS, Kubernetes, Terraform / OpenTofu, and Python. The choice depends on the system and the people who will maintain it.

Data Engineering

I build and evolve data platforms with the teams that produce, process, and use the data. That includes connecting product systems to analytics and ML, agreeing on data contracts and ownership, and making the platform usable and operable. My experience includes Kafka, RabbitMQ, Spark, Databricks, and Dagster.

Technical Leadership

Sometimes a team needs someone to make technical decisions, coordinate with other teams, or establish who owns what. I can take that on as an interim lead or work alongside the existing lead. That has included hiring, mentoring, incident management, and helping engineers take responsibility for their systems.

Selected Project Areas

High-performance distributed systems

I built systems for real-time bidding at E4E5 and micropayment processing at Smart Processing using Erlang. At Smart Processing, this included the micropayment processing core.

Event Tracking & Streaming Platforms

At FREE NOW, I worked on the move to managed Kafka infrastructure and built streaming applications for fraud detection and pricing validation. The work included GPS data, ML inference, schema maintenance, and the operational details of running those services.

At Exmox, I helped evolve event tracking and data infrastructure with Product, Analytics, and ML teams, including platform ownership and engineering practices.

Data Platforms

I’ve built and evolved data platforms across FREE NOW, Adtriba, and Exmox. The work combines architecture and implementation with bringing engineering, analytics, and data science teams together around shared systems.

At FREE NOW, that included managed streaming infrastructure and a cross-functional team building fraud detection and pricing validation. At Adtriba, I built the data platform and team, enabled analytics and data science workflows, and established operational practices. At Exmox, I worked with Product, Analytics, and ML on event infrastructure, platform ownership, and engineering standards.

The platform includes the interfaces between teams: who produces the data, who depends on it, who owns a failure, and how changes are agreed upon.

Infrastructure automation for WSI software

At Mindpeak, I automated AWS infrastructure for whole-slide imaging (WSI) software, including GPU autoscaling, workload orchestration, and Python services.

Open Source

I made kafkus to make Kafka easier to explore during development and debugging.

Working Style

I want to understand the problem before choosing the solution. I prefer changes that people can understand, operate, and revise later.

That means looking at what is already working, being explicit about trade-offs, and leaving the team able to maintain what we build. It also means questioning the architecture when the original reasons for it no longer apply.

Contact

Tell me what you’re trying to do, what is getting in the way, and what sort of help you need. We can work out whether I’m useful.

Get in touch on LinkedIn · GitHub