What each interview
actually covers
Five roles, chosen from what people actually practise here rather than from search volume. Every figure on these pages comes from the same 177 graded interviews by 71 people that the published rubric draws on, between 26 June 2026 and 28 August 2026 — the role counts below are a partition of that sample, not a separate one. Rubric figures as of 28 August 2026; role breakdown as of 30 August 2026.
- AI / Machine Learning Engineer interview38 graded interviews · 10 people · 6 formatsWhat it covers
- Software Engineer interview37 graded interviews · 30 people · 8 formatsWhat it covers
- Data Engineer interview28 graded interviews · 6 people · 5 formatsWhat it covers
- Full Stack Engineer interview22 graded interviews · 13 people · 9 formatsWhat it covers
- Python Backend Engineer interview16 graded interviews · 15 people · 3 formatsWhat it covers
How these five were chosen
The five roles below are the five largest role families in our own graded-interview data, not a list we thought would rank. They were chosen by running the query above against production and taking the top five by graded interviews, with distinct candidates as the tie-breaker so one heavy user cannot manufacture a role. Roles people do not actually practise here do not get a page, however much search volume they have.
The five role pages account for 141 of the 177 graded interviews in the published sample; the remaining 36 are spread across roles too thin to say anything useful about. We would rather publish five pages with a denominator than fifty without one. Rubric figures as of 28 August 2026.
Read the full scoring rubric for what the dimensions mean and where the score is weakest, and our evaluation of the scorer for how well it actually separates candidates.
The query behind every figure on these pages
SELECT lower(btrim(candidate_info->>'role')), count(*), count(DISTINCT customer_id), count(*) FILTER (WHERE recommendation='YES'), count(*) FILTER (WHERE recommendation='MAYBE'), count(*) FILTER (WHERE recommendation='NO') FROM ai_bot_interview_sessions WHERE status='COMPLETED' AND feedback ? 'overallEvaluation' GROUP BY 1