Learn AI through gamesLearn how to prompt AI through exercises
Hands-on prompt engineering exercises grounded in real B2B scenarios: sales, marketing, ops, finance, product. Built for the people on your team who use AI every day and aren't getting much out of it yet.
First 3 exercises free. No signup needed to start.
The problem
Most AI training is watched. Ours gets used.
Hour-long video courses get 14% completion on average. Our bite-sized, hands-on exercises get finished because they're built like a game: short loops, instant feedback, real scenarios from the work your team is already doing.
Hands-on, not theoretical
Every exercise is, well, an exercise: write a prompt, get scored, see the model answer, understand why it works. No lectures, no buzzwords.
Built on real research
The curriculum maps to techniques validated across thousands of papers and real-world deployments. We bust the popular myths (role prompting, threats, magic phrases) and teach what actually moves the numbers.
Made for teams
Admin dashboard, completion tracking, certificates, leaderboards, SSO. L&D gets the rollup. Managers see who's progressing. Your team gets something they'll actually finish.
How it works
Three steps from kickoff to results
Set up your team
Invite people by email or connect SSO. Assign seats, group by department, and decide which modules to make required.
Your team works through the modules
Four modules: Clarity, Context, Reasoning, Agents. Each one is a series of 10-minute exercises using real B2B scenarios from sales, marketing, ops, and product.
Track progress, earn certificates
Admins see completion rates and time-on-task. Each module ends with a certificate your team can post to LinkedIn. Friendly leaderboards keep momentum up across the org.
The curriculum
Four modules, built in order
Start with Clarity. Each module builds on the one before it, from writing a single strong prompt to building production-grade AI workflows.
Clarity
Specific prompts beat clever ones.
The eleven foundational techniques every working professional should know: specificity, examples (few-shot), output formatting, decomposition, self-criticism, constraints, additional information, and how to spot the popular techniques that don't actually work.
Context
Background information is where most prompts come alive.
Constraints, audience modeling, grounding in source material, and the difference between context that helps and context that hurts. Teach the model what it does not yet know.
Reasoning
Get the model to think out loud.
Chain-of-thought, decomposition, self-critique, and ensembling. The techniques that turn a confident wrong answer into a slow, correct one.
Agents
When the model stops talking and starts acting.
Tool use, multi-step plans, evaluation, and the security problems that come with letting a model touch real systems. Build prompts that drive coordinated action without breaking things.
From 'we should use AI more' to 'we use AI well.'
Your team learns it the way they learn anything that matters: by doing it, getting feedback, fixing what didn't work, and moving on to the next exercise. One short rep at a time.
First 3 exercises free. No credit card. No login required.