Coding
60mApplied programming against a realistic problem, with emphasis on working code and reasoning.
- Problem solving
- Code quality
- Testing instinct
Coverage
Every interview carries its own rubric
A coding exercise and a system design discussion are not the same assessment wearing different names, so they do not share a rubric.
Coverage
Interview types we support.
Each type carries its own rubric and competency set. Criteria can be adjusted per role.
Algorithms
60mComplexity reasoning and algorithmic decomposition on problems with more than one viable approach.
- Complexity analysis
- Decomposition
Data Structures
60mSelection and trade-offs of core data structures under stated constraints.
- Trade-off reasoning
- Fundamentals
System Design
75mRequirements clarification, data modelling, failure modes and explicit trade-offs at scale.
- Architecture
- Scale
- Failure modes
Backend
60mService design, data access patterns, API contracts and operational concerns.
- API design
- Data modelling
Frontend
60mComponent architecture, state management, accessibility and rendering performance.
- Component design
- Accessibility
Mobile
60mPlatform fundamentals, lifecycle handling, offline behaviour and release considerations.
- Platform depth
- Lifecycle
Cloud
60mManaged service selection, networking, cost awareness and operational trade-offs.
- Service selection
- Cost awareness
Infrastructure
60mCI/CD, observability, incident response and infrastructure-as-code practice.
- Observability
- Incident handling
Data Engineering
60mPipeline design, data quality, orchestration and schema evolution.
- Pipeline design
- Data quality
Machine Learning
75mProblem framing, evaluation design, data realities and production considerations.
- Problem framing
- Evaluation design
AI-Driven Development
60mBuilding with AI coding tools as part of the workflow: prompting, reviewing generated code, and judging when to trust it versus rewrite it.
- Tool judgement
- Code review discipline
Product Management
60mPrioritisation under constraint, stakeholder trade-offs, and turning ambiguous problems into a shippable plan.
- Prioritisation
- Stakeholder judgement
Behavioural
45mStructured situational assessment of collaboration, ownership and communication.
- Collaboration
- Ownership
Technical Leadership
60mTechnical direction, cross-team influence and handling of ambiguity at scale.
- Influence
- Judgement
Engineering Management
60mPeople development, delivery management and organisational decision-making.
- People development
- Delivery
Scoring
One scale, and everybody knows what it means.
A 4 has to mean the same thing on every interview or cross-interview comparison is nonsense. The scale is platform-wide; individual criteria add behavioural anchors describing what each level looks like for that specific question.
- 1
Strong concern
Well below the bar for this level.
- 2
Below expectations
Did not consistently meet the bar.
- 3
Meets expectations
Met the bar for this level.
- 4
Above expectations
Exceeded the bar with clear evidence.
- 5
Exceptional
Well above the bar; rare in practice.
What an anchor looks like
Asking somebody to rate “scalability” out of five without saying what a three looks like invites them to invent their own scale, and two interviewers with two private scales are not comparable. So criteria carry anchors. For identifies bottlenecks in a system design interview:
- 1 — Did not identify where the design would break first.
- 3 — Found the obvious bottleneck and addressed it.
- 5 — Found a non-obvious one, and was right about it.
Evidence is not optional
On the criteria that matter most, a rating on its own will not submit. The interviewer has to write down what they observed — what the candidate said or did, not how it felt. That field is the one your hiring team will actually read, and it is the reason an evaluation survives a debrief three weeks later.
Your rubric, if you want it
Every type ships with a rubric written by our team, and you can adjust the criteria for your own roles. What you cannot do is change a rubric for an interview already under way: the evaluation is stored against the rubric it was collected under, so an assessment written in January still shows January’s questions.
Two ways to start.
Hiring teams can walk through their requirements with us. Experienced engineers can join the interviewer network.