Interview rubrics
How Tech Interviews Are Scored: Rubrics by Round and Level
Tech interview rounds are scored dimension by dimension, each rated 1 to 4 against anchors that change with the level you are interviewing for. A clean answer can be outstanding at mid level and borderline at senior level, because senior candidates are expected to raise the important points without being asked. Below are the 13 rubrics our AI interviewer scores against, one per round type.
Rubrics by round
- Behavioural interview rubric
6 dimensions; heaviest is situation and ownership (20%).
- Coding interview rubric
5 dimensions; heaviest is approach and optimisation (25%).
- Frontend Coding interview rubric
5 dimensions; heaviest is language and platform fluency (25%).
- Communication Assessment interview rubric
5 dimensions; heaviest is fluency (20%).
- HR interview rubric
4 dimensions; heaviest is motivation and fit (30%).
- Low-Level Design interview rubric
6 dimensions; heaviest is class and entity modelling (25%).
- ML System Design interview rubric
6 dimensions; heaviest is problem framing (20%).
- Online Assessment interview rubric
3 dimensions; heaviest is correctness (tests passed) (50%).
- Project Deep Dive interview rubric
6 dimensions; heaviest is technical depth (25%).
- SQL interview rubric
6 dimensions; heaviest is query correctness (30%).
- System Design interview rubric
6 dimensions; heaviest is requirements and scoping (20%).
- CS Fundamentals interview rubric
4 dimensions; heaviest is concept accuracy (30%).
- Troubleshooting interview rubric
6 dimensions; heaviest is systematic investigation (25%).
Company-specific signals
Where this rubric comes from
These rubrics are researched: they follow the scoring methods big-tech companies describe publicly (behaviourally anchored ratings, bar raisers, leadership principles, named interview signals) and how their rounds are commonly run. No company publishes its actual scoring sheet, so the level-by-level wording is our own synthesis. It is the same rubric our AI interviewer scores every mock against.