Principal Software Engineer job description

Software engineering — Principal
Shapes engineering strategy and the hardest calls org-wide.

REQUIRED SKILLS
Software Development
- Programming and Build — You can hold technical responsibility for every stage of a software project spanning multiple teams, and you advise the organisation on which standards, languages and tooling to adopt or retire. You are asked to judge the highest-risk changes before they ship.
- Modern Development Standards — You can decide which development standards the organisation adopts, pilots or retires, and show evidence of the effect on delivery. You change practice in teams you do not manage by persuasion rather than mandate.
- Code Review — You can shape how review works across the organisation, including what must be reviewed, by whom, and where review should be replaced by a stronger control. You are called in on the changes nobody else feels able to approve.
- Debugging and Diagnosis — You can diagnose failures that cross system and organisational boundaries, and you drive the changes that make whole classes of defect either diagnosable or impossible. Teams escalate to you as the last resort.
- Refactoring — You can set how the organisation pays down structural debt, including which systems get invested in and which get replaced or deliberately left alone. You are trusted to call an expensive rewrite off.
- Version Control — You can define the branching and release workflow your team uses and explain the tradeoffs it makes. You handle the situations others avoid, such as a rewritten shared branch or a secret committed to history.
- API Design — You can set API standards for the organisation, including how long compatibility guarantees hold, and arbitrate between teams with conflicting interface needs. You design the interfaces that outlive their systems.

Testing and Quality
- Test Design — You can set organisational standards for test types, techniques and coverage, and show which risk each one addresses. You drive continuous refinement of how testing is done rather than defending the current suite.
- Test Automation — You can set the automation strategy for a product, deciding what runs where, on what trigger, and what blocks a release, and you evaluate tooling against real needs rather than novelty.
- Defect Management — You can lead improvement of defect and test processes, help others assess complex or contested defects, and use defect patterns to make specific recommendations about where to invest. You coach others in doing the same.
- Test Strategy — You can own the quality strategy for the organisation and advocate for testing to be engaged at the point where it is cheapest, holding that line when delivery pressure makes it unpopular.
- Performance Testing — You can set how the organisation tests and budgets for performance, including what is measured in production rather than in a lab, and influence architecture decisions before performance has to be retrofitted.
- Accessibility Testing — You can set the accessibility bar for a product and the checks that enforce it, champion it with internal stakeholders, and judge which findings are genuine barriers rather than tool noise.

Systems and Design
- Systems Design — You can set the design direction for systems the whole organisation depends on, and you are asked to judge whether a proposed design is worth building at all. You make the case for the simpler option when it is unpopular.
- Systems Integration — You can set how systems across the organisation are connected, including which integration patterns are standard and which are being retired, and you resolve disputes between teams. You can say when two systems should not be joined at all.
- Architectural Decisions — You can set which decisions need recording across the organisation, who takes them and how they are reviewed. You are asked to arbitrate the choices where two teams each have a defensible answer.
- Schema and Query Design — You can set data modelling and query standards across the organisation and decide which stores are used for what. You are the person asked to check the migrations that cannot be undone.
- Prototyping and Spikes — You can use prototypes to take risk out of complex work, and you set how far a spike runs before the team commits to a direction. You are willing to report that the prototype disproved the option you preferred.

Operations and SRE
- Continuous Delivery — You can set delivery practice across the organisation, including what evidence a release needs, and show the effect on lead time and failure rate. You hold the line on small frequent releases when a team wants to batch them.
- Observability and Monitoring — You can set observability standards and tooling across the organisation, including what every service must emit, and bring the cost of that data down without losing what matters. You are asked why a whole class of failure went unseen.
- Production Incident Response — You can set how the organisation responds to incidents, including severity, escalation and what a follow-up must produce. You read the pattern across incidents and use it to argue for the structural work that prevents the next one.
- Availability and Capacity — You can set how the organisation defines, funds and reports reliability, and arbitrate between availability and delivery speed. You say plainly when a target costs more than the risk it removes.
- Infrastructure as Code — You can design the infrastructure codebase for a product, set the patterns other teams reuse and bring manually built resources back under code. You keep a risky change reviewable by making the plan small enough to read.

Security and Data
- Application Security — You can set security engineering practice across the organisation, decide which controls are mandatory and work with security specialists on the risks the product actually carries. You are asked to judge whether a residual risk is acceptable.
- Secure Data Handling — You can set how a product handles sensitive data, work with legal and privacy colleagues on the obligations, and build controls that make compliance provable rather than asserted. You find the data flow nobody wrote down.
- Dependency Management — You can set how a product takes on and retires dependencies, plan an upgrade that touches many services and handle a supply-chain advisory quickly and calmly. You get a stuck legacy dependency moving rather than recording it as permanent.

Delivery
- Project Management — You can run the organisation's largest and most contested programmes, and the planning practices you introduce get adopted by teams you do not lead.
- Planning & Estimation — You can forecast at the scale of quarters and headcount, and the estimation practice you set is what the wider organisation plans against.
- Ownership & Accountability — You can take on the outcomes the organisation is most exposed on, and other leaders route ownerless problems to you by default.
- Quality Focus — You can raise the quality bar across the organisation by building the tooling and habits that make the careful path the easy one.

Craft
- Problem Solving — You can crack problems the organisation has repeatedly failed to solve, and your approach becomes how others tackle that whole class of problem.
- Domain Expertise — You can influence how the field is practised beyond this organisation, and your expertise settles questions that have stood open for years.
- Continuous Learning — You can set what the organisation invests its learning time in, and the material and practice you create outlive your involvement.

Communication
- Communication — You can set how the organisation communicates, and the forums and norms you create measurably improve how information travels through it.
- Collaboration — You can dismantle the barriers that stop teams working together, and how the organisation collaborates changes because of what you built.
- Technical Writing — You can set the writing standard the organisation works to, and documents you authored are still in use long after you moved on.
- Stakeholder Management — You can represent the organisation in its most consequential relationships and set how it engages with everyone who depends on it.

Leadership
- Leadership — You can set direction for the whole organisation, make hard calls under real uncertainty, and carry the decisions nobody else wants to own.
- Mentoring — You can shape how the organisation grows its people, and those you developed are themselves named by others as strong practitioners.
- Strategic Thinking — You can shape the organisation's strategy, and the bets you argued for are visible in where it ended up.