Episode Highlights
- From AI Hype to Practical Application (02:20)
Framing AI within an improvement context, with a clear argument that value comes from disciplined application and experimentation, not adopting technology for its own sake.
- What AI Agents Are and Why Context Matters (08:17)
A combined explanation of AI agents as contextualised, repeatable instructions layered onto AI models, alongside a practical analogy that shows why general capability isn’t enough without clear instructions, constraints and context.
- Agents as a Tool for Standardisation and Reducing Variation (11:28)
Linking directly to core improvement thinking, positioning agents as a way to standardise multi-step tasks and reduce variation where prompts or decisions would otherwise differ between individuals.
- From One-Off Prompts to Repeatable Processes (15:00)
A walkthrough of how prompts can be codified, saved and refined over time, shifting AI from ad hoc experimentation into structured workflows that become more reliable with use.
- Adding Automation to Agents (18:41)
Exploring the step change when agents move beyond generating responses to triggering actions, integrating with tools and workflows to complete tasks end-to-end.
- Applying PDSA to AI: A Structured Way to Experiment (21:08)
Introducing a practical model for adoption, using evaluation, recruitment and testing to ensure AI solutions are grounded in process thinking and measurable improvement.
- Separating “Thinking” from “Doing” in AI Design (24:10)
A useful distinction between AI as a source of intelligence and automation as execution, showing why combining both is essential for building scalable, repeatable solutions.
- From Small Use Cases to Organisational Capability (42:12)
Closing with the idea that even simple, localised use cases build capability that can then be applied more widely, supporting broader transformation across processes and teams.
