Kagento

About Kagento

Kagento is a technical-hiring assessment platform built around a premise most of the category has avoided: candidates use their own AI tools during the assessment. Rather than locking a candidate into a browser editor with autocomplete disabled, Kagento drops them into an isolated Linux container or vcluster with Claude Code, Codex, Cursor or whatever setup they would actually use at work, and records everything. What the hiring team reviews afterwards is not a diff but a session replay: every command, every edit, every test run, in order, so a reviewer can see how the candidate directed the agent, whether they verified its output, and how they recovered when it was wrong. Those are the skills the job now requires and the ones a traditional assessment cannot observe. The company is deliberate about the boundary it will not cross — the site states there is no AI interviewer and no AI hiring decision, and scoring is hidden tests plus human review of the replay rather than a model's judgement of the candidate. The task catalogue covers backend, frontend, platform, Kubernetes, ML and security roles with over a hundred real-world tasks and tamper-resistant task variants. Hosting is EU-only in eu-central-1 and GDPR-compliant by default, with a printable scorecard and JSON export for pushing results into an ATS.

Key Features

Tags