Operational Excellence Through Data
Engineering leaders are asked more and more to show that delivery is healthy and that AI investment is paying off. Dashboards are easy to build and hard to trust. I help teams measure delivery and reliability in ways that survive scrutiny, and then act on what the numbers say.
Who this is for
- Heads of engineering who need to report delivery performance to a board or CEO.
- Teams with metrics tooling in place, but little confidence in what it shows.
- Organisations rethinking incident response and reliability after an outage.
- Leaders who suspect their test suite tells them less than it appears to.
What I do
- Delivery metrics. DORA and related measures, with a clear view of what each can and cannot tell you, including in the presence of AI assistants. See Decoding the 2025 DORA Report.
- Agile measurement. Replacing velocity and estimate accuracy with flow-based measures. See The Agile Metrics That Actually Matter and Why Agile Estimation Fails.
- Operational excellence as a shared responsibility. Reliability as a team habit, not a separate function. See Why Operational Excellence Must Be Everyone’s Responsibility.
- Incident response. Preparing for failure and learning from it. See The New Reality of Incident Response.
- Testing visibility. Knowing what your tests actually do across environments. See The Testing Visibility Gap. I also founded Obvyr, a test observability product, because of this problem.
How it starts
A free introductory call of 15 to 30 minutes, then, where it fits, a fixed-price audit of one to two weeks. See the engagements page.
Questions
Are DORA metrics still useful with AI assistants?
They remain a useful starting point, but they were not designed to measure AI contribution. I read them alongside review quality and the evidence behind any productivity claim.
Can you help us choose a metrics tool?
I can help you decide what you need to measure and why before you choose one. Tooling decisions are easier once that is clear.
Do you work with our existing tools?
Yes. The first step is understanding what you already collect and how far it can be trusted.
Related
Browse writing on delivery metrics and testing visibility, or see AI transformation.