A 10-minute hands-on check shows how people actually work with AI: whether they get the right answer, catch its mistakes, and know when not to trust it. Not a quiz about prompts.
What we measure
Every result comes from what someone actually did with an AI assistant on a realistic task, scored by fixed rules.
Did the work that would actually be sent come out right?
Did they check the AI's output, catch the planted error, and ask it to show its work?
Do they know when to trust AI, which tool to use for company data, and how to give it context?
A 3-minute survey: size, functions, AI tools you pay for, your goals.
10 minutes, so you know exactly what your team will see.
Post it in Slack. No accounts, no installs.
Unlocks at 5 people: where AI output goes unchecked, confidence vs. reality, by function.
A 90-minute assessment on real work for the functions that need it, with evidence and a practice plan.
Private by design
When 10 minutes isn't enough
Each person completes a real task from their function with an AI assistant. Every finding in the report links to the evidence. A reviewer signs off. Then targeted practice and an unseen retest show whether it stuck.
Team pilot: $99 for 10 people
Benchmark my team →No. You do a short piece of real work with a built-in AI assistant. We look at the result you'd actually send and how you got there.
By fixed rules: whether your answer is right, whether you caught the error planted in the material, and your choices on four judgment questions. No AI grades you.
Only you, unless you choose to share it. Team leads see aggregates once 5 or more people have finished.
The 10-minute check and the team report are free. Deep assessments for teams are paid.