Utkrusht AI

Utkrusht is a hiring assessment platform for tech teams that want to evaluate candidates by watching them solve real work instead of relying on quizzes, take-home assignments, or interview guesswork. It is built around “watch-them-work” tasks, where candidates solve, debug, design, extend, or refactor in a live production environment while the session is recorded and scored.

The platform is aimed at small and mid-sized companies hiring technical roles such as fullstack, SRE, AI engineer, DevOps, data engineer, and backend. It helps hiring teams understand not just whether someone can write code, but how they think, make tradeoffs, debug messy systems, explain their approach, and use AI tools during the work. According to the page, this is used to generate a ranked shortlist of the top candidates so interviewers can spend time only on the most promising people.

A typical workflow starts by creating a position or uploading a job description. The platform extracts the role, matches relevant skills, and builds a live task from a library of customizable templates. Candidates then complete the assessment, and the platform returns detailed analysis, scores, video session recordings, and a shortlist. It also supports screening signals such as intent to join, location match, salary alignment, and culture fit, and it tracks AI usage and cheating during sessions.

The product emphasizes quick setup, a 30-minute assessment format, and candidate completion without requiring a credit card to start a trial.

Key features:
- Live production-environment assessments instead of quizzes or artificial simulations.
- Automatic task generation from a job description or customizable job templates.
- Ranked shortlists based on technical execution, AI usage, communication, judgment, and problem-solving.
- Video-recorded sessions with detailed analysis and score breakdowns.
- Signals for candidate rubric factors such as intent to join, location match, salary alignment, and culture fit.
- Proctoring and session review to detect cheating and show how AI tools were used.
- Designed to help teams review only the top 5–10 candidates after screening.