Profile
I work across the whole path a change takes: the code, the review, the build, the deploy and the dashboard someone looks at when it goes wrong. Most of my contracts start with one of those being slow or fragile, and end with all of them written down.
Alongside client work I test AI tools, agents and self-hosted software on my own hardware and write up what they cost, what licence they carry and where they fall short. Those reviews are the best sample of how I think and write.
Sample text: the experience entries below describe kinds of work, not named clients or dates, until Lance’s own CV is supplied.
Skills and tools
Application code
TypeScript, front end and back end
Node.js services and APIs
Python for data jobs and tooling
PostgreSQL design and tuning
Delivery and infrastructure
Docker images and Compose
Kubernetes clusters and Helm
CI/CD on GitHub Actions
Ansible for servers outside a cluster
Operations
Grafana dashboards and alerts
AWS accounts and cost reviews
AI tooling
Self-hosted models with Ollama- Agent pipelines and model cost tracking
Experience summary
-
Contract, ongoing
Full-stack and DevOps contractor
Web applications in TypeScript and Python, delivery pipelines, Kubernetes platforms and cloud cost work, for teams that need one person across all of it.
-
Contract work
Platform and pipeline rebuilds
Moving slow or hand-run deploys to CI with tested rollbacks, container images and infrastructure written as code.
-
Contract work
AI tools brought in-house
Choosing, hosting and measuring models and agents where data has to stay inside the company, with the running cost set out before the build.
-
Ongoing
Writer, lancekeay.com
Hands-on reviews of AI tools and self-hosted software, and a monthly report comparing AI model token prices.