Kendor Reviews and Details

Questions & Answers

1. What makes Kendor unique?

   Kendor runs technical interviews in a real development environment, not a puzzle editor. Candidates work in a full VS Code editor in the browser, backed by a secure cloud sandbox that runs their code. That can be your own repository, with real services, a live preview and databases like PostgreSQL and Valkey. They can use AI coding agents such as Claude Code and Codex when you allow it, and reviewers see exactly how the AI was used. Live interviews happen in the same environment, with a built-in video call. You see how someone actually works, not how well they memorised algorithm puzzles.

2. Why should a person choose Kendor over its competitors?

   Most assessment tools still test isolated puzzles in a simple editor, and treat AI as cheating. Kendor reflects how engineers work today:

   • Real projects instead of puzzles: start from your own GitHub or GitLab repository or a ready template, in Node.js, TypeScript, Python, Go, Java, Rust, C++ or PHP.
   • AI allowed and visible: you decide whether candidates may use AI agents, and every prompt and change is there for review.
   • One place for take-homes and live interviews: the same environment, with video calls, a shared editor and review summaries.
   • Fits your hiring process: integrations with Ashby, Greenhouse and Lever. Candidates need no account and nothing to install.
   • Clear pricing: from $79/month with unlimited candidates, with an annual discount, instead of opaque per-seat enterprise quotes.
   • Built for the EU: GDPR-ready, with data stored in the EU.

3. How would you describe the primary audience of Kendor?

   Engineering-led hiring teams at startups and growing tech companies: CTOs, engineering managers and the recruiters who work with them. They want to see how a candidate really builds software, including with AI tools, rather than how they score on algorithm puzzles. Kendor suits teams hiring a few engineers at a time as well as companies running technical screens at volume.

4. Which are the primary technologies used for building Kendor?

   TypeScript throughout:

   • React with TanStack Start for the web app
   • NestJS with PostgreSQL (Prisma) and Valkey for the API
   • a Docker-based sandbox service, built with Fastify, that runs each candidate’s code in an isolated container
   • VS Code’s editor running in the browser, with language servers inside the sandbox
   • LiveKit for video calls and Yjs for real-time collaborative editing
   • integrations with Anthropic (Claude Code) and OpenAI (Codex) for AI agents
   • hosted in the EU, with storage on Amazon S3 and delivery through Cloudflare

5. What's the story behind Kendor?

   I started building it when I was hiring engineers and found no good solution in the market. Most at the time focus on Leetcode like puzzle-tests and I didn't think they were realistic skill assessment for the job. So for the first couple of hires, we were just sending candidates repos with READMEs as take home assessment with manual timer. But the problem with that is we have no insights to how they work or what's their thinking process was while implementing it(Now with Agentic coding, it's even harder). And everything was manual including reviewing process so it wasted a lot of time. So I built Kendor to make my own hiring process easier and my colleagues liked it as well so they started using it for hiring as well.