Four disciplines. Not a list of skills — a way of working.
Each one links out to the actual projects, articles and talks behind it.Each discipline is supported by practical projects, frameworks and published work—not simply claims of expertise.
Incident Engineering
Treating a production failure as a subject to be investigated properly, not a fire to be put out and forgotten. That means reconstructing what actually happened, reviewing the decisions made in the moment, and turning both into something the team can act on.
- Production failures
- Incident reviews
- Operational decision-making
Forward Deployed Engineering
Getting close enough to a customer's actual environment to see the problem they can't fully describe. That's technical investigation done in context, alongside the people who live with the system every day — not from a support queue.
- Customer problem solving
- Technical investigation
- Cross-functional collaboration
AI-assisted Operational Judgement
Using AI to support investigation — surfacing patterns, summarising signals, speeding up the boring parts of an incident — without letting it replace the human judgement call. The model helps you decide faster. It doesn't decide for you.
Reliability Engineering
The upstream work that makes incidents smaller and rarer in the first place — knowing the failure patterns a system is prone to, testing operational readiness before it's needed, and treating reliability as something you keep improving rather than a box you tick once.
- Failure patterns
- Operational readiness
- Continuous improvement