Venture Scout
Turn five conflicting sources with around 250 startup entries into a reliable shortlist, accounting for duplicates, outdated details and marketing claims.
Interactive AI learning
Your team learns AI through decisions, not slides.
AI Quest turns reliable AI use into realistic missions. Participants experiment, receive immediate feedback and apply what they learn to their everyday work.
AI Quest · Mission: Invoice check
Product concept, schematic · Level 3
Pipeline of building blocks
- RuleCheck duplicates & unknown IBANs
- AIExtract fields from PDFs
- RuleCompare with purchase order
- RuleAdd uncertain items to the query list
- HumanApprove using the approval matrix
Assessment across six dimensions
- Output quality30%
- Robustness20%
- Human Control20%
- Efficiency15%
- Traceability10%
- Resource use5%
How a mission works
Every mission uses only curated synthetic practice data, including realistic challenges: duplicates, missing values, conflicting sources and deliberate edge cases where blind trust in AI fails.
01
Experience and measure the manual task
Participants first complete the task themselves and see where time and quality are lost.
02
Build your own pipeline
Build a workflow from clear rules, AI steps and human approvals, then test it against edge cases.
03
Feedback across six dimensions
See how the workflow performs on result quality, robustness, human control, efficiency, traceability and resource use.
Example missions from the product
Invoice Check
Turn PDFs into verified accounting entries: AI extracts the data, rules match it against purchase orders, and people approve it according to the approval matrix.
Market Intelligence
Monitor the market with source evaluation: primary sources before trade publications before blogs, with clear limits on what AI can substantiate.
Six levels to the Control Tower
Higher levels are unlocked by demonstrating competence, not by consuming content.
- L0Manual RealityExperience process friction and sources of error yourself
- L1Prompt PartnerDefine the goal, context and quality criteria for an AI task
- L2Grounded DataSelect sources and check data quality
- L3Workflow & LoopsBuild rules, exception handling and tests for repeatable workflows
- L4Agent + ToolsSet permissions, budgets and stop rules for agents
- L5Multi-Agent Control TowerCoordinate specialist roles and approve actions that have real effects
What participants practise
- Assign tasks appropriately to rules, AI and people.
- Make sources, data quality and uncertainty visible.
- Set limits on automation before anything is sent, booked or approved.
- Assess results across several dimensions of quality.
- Apply what you have learned to one of your own work processes.
Who AI Quest is for
- Leadership teams looking to build a shared understanding.
- Business teams preparing to start an AI project.
- Learning and enablement programmes with a practical focus.
- Companies that want to explore opportunities and concerns constructively.
From the game to everyday work
AI Quest has a practical purpose. Its lessons provide a useful starting point for prioritising real processes, clarifying responsibilities and implementing a first workflow with appropriate controls.