Programme · 42 modules
1. Understand
FreeWhat AI can and can't do
FreeThe 4 families of AI
SoonEssential vocabulary
SoonMyths that sink projects
2. Frame
FreeDo you really need AI?
SoonDefine the problem
SoonScope and out of scope
SoonMeasurable success criteria
SoonStakeholders and RACI
SoonThe 1-page project brief
3. Data
SoonData inventory and quality
SoonOwnership, licences, rights
SoonPersonal data and GDPR
SoonTest and evaluation sets
4. Choose
SoonBuy, assemble or build
SoonChoosing a model
SoonOpen source vs proprietary API
FreeSolid prompting in 5 blocks
SoonContext, RAG and memory
SoonAgents and guardrails
5. Prototype
SoonThe 48-hour prototype
SoonIterate with test sets
SoonMeasure accuracy, cost, time
SoonReview and go / no-go
6. Quality and ethics
FreeHallucinations: reduce them
SoonBias and fairness
SoonSecurity: injection, leaks
SoonCompliance: AI Act, GDPR
SoonHuman in the loop
7. Deploy
SoonMinimal architecture
SoonIntegrate with existing tools
SoonChange management
SoonBudget and cost caps