Hire for AI fluency, in every role.
Every role is a technical role now. Give candidates a real assignment, real files, and real models, and watch whether AI makes them faster and better.
Try it — click Files on the left to switch the output, or prompt AI Assist on the right.
Campaign Brief & Interview Task
Scenario
You’re a growth marketer at a B2B SaaS company. Last quarter’s multi-channel campaign just wrapped, and leadership wants a clear read on what worked before the next budget cycle.
Your task
Using the attached data and AI Assist, produce a summary leadership can act on. Specifically:
- Pull the key metrics from the channel spend data.
- Build a simple HTML dashboard that highlights performance by channel.
- Recommend where to shift budget next quarter, with reasoning.
What we’re assessing
- Prompt quality and specificity
- Iteration and error-checking
- Judgment applying AI output to a real decision
The messy files the job actually hands them
Candidates don’t get emails. They get a workspace with the same materials for the role. You upload the file or use our templates.
Because everyone gets the same set, you’re comparing candidates on the same criteria — and you find out fast who can perform on the job to solve a realistic work challenge.
Watch real prompting, iteration, and judgment
Every candidate gets a consistent experience with Claude, GPT, Gemini, and Llama built into the pad. Candidates don’t bring their own accounts, and your team doesn’t configure anything. You see the candidate, not their tooling.
Replay every prompt, edit, and AI response
Every prompt, edit, model switch, and response is captured — scrub the timeline to the moment the work turned, or skip straight to the usage summary and make the call in two minutes.