AI in Action
A hands-on workshop where your team uses real data and AI to solve real problems, in real time. No slides. No theory. Just results.
What you'll
walk away with.
This session is designed to show your team what AI can actually do when it's pointed at real data. By the end, every person in the room will have built something with AI that would normally take hours, in under 15 minutes.
Practical AI confidence
Every team member will have personally used AI to analyse data and generate strategic output. Not watched a demo. Done it themselves.
A repeatable framework
The RICCE prompting method, Role, Intent, Context, Constraints, Examples, that turns vague AI requests into precise, usable outputs every time.
Two working prototypes
Real notebooks your team can take away, adapt to client data, and use in production workflows immediately.
Two hours,
four blocks.
A rhythm that moves from foundations to advanced techniques to hands-on to debrief, so the learning actually sticks.
Foundations
- What AI actually is (and isn't): mental model reset
- The RICCE framework for prompting
- Live demo: messy brief to polished output in 60 seconds
- Common mistakes that make AI useless, and how to avoid them
Advanced AI Frameworks
- Prompt chaining: building multi-step workflows that compound in quality
- Few-shot and zero-shot techniques for consistent, structured outputs
- System prompts and role-setting for domain-specific analysis
- When to constrain AI vs. when to let it explore
Hands-on · Choose one
Task A or Task B
- Option A: The AI Greenlight Committee, interrogate 40,000 movies to greenlight or pass on a pitch
- Option B: AI Marketing Strategist, reverse-engineer why a film worked or flopped and build a campaign brief
- Pick the task that fits your role, then present your output to the room
Debrief & Q&A
- Team share-back: what surprised you?
- Where does this fit in your current workflows?
- Open Q&A
Pick your task.
Real data.
Choose the challenge that fits your role. Each person, in their own browser, using the same dataset the whole room is looking at. No spectators.
The AI Greenlight
Committee
Scenario
You're a studio executive. A producer has just walked into your office with a pitch for a new film. Your job: use AI to interrogate a dataset of 40,000 movies, budgets, revenues, ratings, genres, and decide whether to greenlight the project or pass.
What this teaches
- How to feed AI structured data and get strategic analysis back
- How prompt specificity changes output quality dramatically
- The difference between asking AI a question and giving AI a job
What you'll do
- Load a real movie performance dataset into Google Colab
- Invent a movie pitch, genre, budget range, concept
- Prompt AI to analyse what historically works at that budget and genre
- Get a greenlight/pass recommendation with projected ROI
- Let the AI play devil's advocate, it challenges your pitch with data
// Marking criteria
AI Marketing
Strategist
Scenario
Pick any movie from the dataset. Your job: use AI to reverse-engineer why it succeeded or flopped, then generate a full marketing campaign brief, as if you were the agency hired to launch it.
What this teaches
- How to use AI for strategic analysis, not just content generation
- How to chain prompts: analysis, diagnosis, creative output
- How AI can accelerate the strategy-to-brief pipeline your team runs every day
What you'll do
- Choose a movie: big hit, hidden gem, or famous flop
- Prompt AI to analyse performance: budget efficiency, genre trends, competition, timing
- AI generates a diagnosis of what worked, what didn't, and what the data says
- Prompt AI for a complete campaign brief: positioning, audience, taglines, channels, timeline
- Present your brief, could a client buy this?
// Marking criteria
40,000 movies.
Real numbers. Real patterns.
Loaded straight into Colab. No setup.
Budgets, revenues, ratings, genres, release dates, production companies and audience scores for roughly 40,000 films, sourced from a public movie database.
You'll use AI to find patterns no human could spot manually in 15 minutes: which genres over-perform at low budgets, when timing kills a good film, and what the data really says about sequels vs. originals.
Download the dataset ↓Practitioners first.
Always.
We Are Pioneers is an AI-first consultancy that trains agency teams to actually use AI, not just talk about it.
We've trained teams at LEGO, Adidas, and the Press Association. Every session is built bespoke. Every exercise uses real data. Every outcome is something your team can use the next day.
// Credentials
Questions after
the session?
Drop Jeremy a message if you need anything after the day. Otherwise, use your curiosity, and be ready to build.
Get in touch →