Advanced level · Module 1 of 6 · 22 guided hours
Designing AI Systems
Decide where AI adds value, design system architecture with humans in the loop and choose models for real-world constraints.
Learning objectives
- Decide whether and where AI should be used
- Design AI system architectures with human oversight
- Choose models and deployment for cost, privacy and connectivity constraints
Topics
- problem framing and AI suitability
- system components
- human-in-the-loop design
- model selection criteria
- data location and privacy
- offline and low-bandwidth design
Lessons
-
Lesson 1 Should We Use AI for This?
Judge whether AI is the right tool for a task and where in a process it should sit.
task analysiserror tolerancealternatives to AIvalue and risk
-
Lesson 2 Architecture and Human Oversight
Design the components of an AI system and place human checkpoints where they matter.
components (interface, retrieval, model, tools, logs)human in the loopapproval stepsaudit trail
-
Lesson 3 Choosing Models
Choose models by task quality, cost, speed, data handling and availability.
hosted vs open-weight modelsmodel size and qualitydata handling and locationvendor dependence
-
Lesson 4 Designing for Low Connectivity
Design AI systems that remain useful with intermittent internet, power and limited devices.
intermittent connectivityoffline fallbackqueueing and syncsmall local models
Practical activity
Produce a design document for the capstone system: user journey, architecture diagram, human checkpoints, model choice and fallback when connectivity fails.
Expected outcomes
- Well-justified AI system designs
- Designs that work in real local conditions
Assessment
Design review against a checklist covering suitability, oversight, constraints and risks.
Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.