Level 3 of 3
Advanced Level
Prerequisites
AI Intermediate level, or equivalent: calling model APIs from Python, prompt engineering, building and evaluating a retrieval-augmented assistant.
Duration
12 weeks · 140 guided hours · 6 modules. Taught live by an instructor — online or in person.
By the end of this level you can
- Design AI systems that fit users, constraints and risks, including low-connectivity settings
- Build agents that use tools safely within clear permissions
- Choose between prompting, retrieval, fine-tuning and open models
- Secure, test and monitor AI systems against misuse and failure
- Govern AI in organisations with policies, impact assessments and value measurement
Modules
- 1Module 1 · 22 h · 4 lessonsDesigning AI Systems
- 2Module 2 · 22 h · 4 lessonsAI Agents and Tool Use
- 3Module 3 · 22 h · 4 lessonsOpen Models and Customisation
- 4Module 4 · 22 h · 4 lessonsAI Safety, Security and Testing
- 5Module 5 · 22 h · 4 lessonsAI Governance and Strategy
- 6Module 6 · 30 h · 5 lessonsAI Capstone Project
Advanced level project
AI Assistant System with Governance Pack
Design, build and evaluate an AI system for a realistic organisational need — for example a case-worker assistant that drafts referral summaries from approved templates and searches policies — together with the governance pack an organisation would need to adopt it.
Deliverables
- A working system using retrieval and at least one tool, with permissions and logging
- An evaluation and red-team report with fixes
- An AI impact assessment, use policy and monitoring plan
- A presentation to 'leadership' with a go/no-go recommendation
Assessed on
- The system design matches user needs, constraints and risks
- Tool use is limited by permissions and sensitive actions require human approval
- Testing covers accuracy, prompt injection and misuse, with evidence of fixes
- Governance documents are complete, realistic and consistent with the system
All levels are taught live by an instructor — online or in person — see registration options.