Backpressure
Practice system-design interviews without another person in the room: diagram an architecture on a canvas, get grilled by an AI interviewer, and read back a scored "would you pass?" verdict.
System-design interviews are hard to rehearse alone. The value is in being pushed, "what happens when this queue backs up, where's your single point of failure, how does this read path scale," and a static prompt list can't push back. So Backpressure puts an interviewer on the other side of the whiteboard: you draw, it interrogates, and at the end it grades.
You start from a prompt, "design Slack," "design a URL shortener," sketch your components and their connections on an open canvas, and talk through the design with an AI interviewer that probes the weak spots. When you're done it returns a scored verdict against a rubric written for that specific problem, judged on the properties your design achieves rather than whether it matches one reference answer, so a genuinely different-but-sound design still passes.
How it works
- An open canvas, not preset blocks: you draw and name every component yourself; the node labels are the whole signal, so the tool never boxes you into one shape of answer.
- An interviewer that pushes back: a chat model interrogates your design in real time, and a second model returns a structured, property-based verdict with a score and the reasoning behind it.
- Graded on properties, not resemblance: each problem's rubric rewards what the design achieves (availability, scaling, failure handling), so a completely alternative architecture that holds up is marked correct.
- ~50 problems: Slack, YouTube, Uber, an LLM inference service and more, each with its own rubric and reference solution.
- Bring your own Claude: it runs on your own Claude Code login; there's no API key to paste, no account to make, and nothing to pay.
Stack
Next.js · TypeScript · Tailwind · @xyflow/react canvas · Claude, through the Agent SDK or the raw Messages API. Free and open source; there's no hosted demo, you clone the repo and run it against your own Claude login.
Next more problems, and calibration so every rubric grades a good alternative design as fairly as the reference one.