How Rewst Uses AI to Raise the Bar on Course Quality with Clea Mahoney

What does it actually take to build an AI workflow that holds the bar on quality? Not the theory — the real thing, with all the context window limitations, mid-project pivots, and "why are you failing me, Claude?" moments.Clea Mahoney is a Learning Experience Design Manager at Rewst, a workflow automation platform for managed service providers, and the CE challenge there is one of the most complex I’ve come across. Every single customer is building something different, which means there's no standard, no one-size-fits-all. She has to figure that out fresh every time.What makes this conversation so good is that Clea isn't talking about AI in theory. She walks us through the actual workflow she's built inside Claude to audit, update, and maintain her course library — from the Notion drafts to the instructional design rubric she trained Claude on, to the multi-chat structure she uses to keep context windows from getting overwhelmed. It’s specific, it’s practical, and it’s the most honest account you'll hear of what it actually takes to build an AI workflow that holds the bar on quality.In this episode, we get into:The course audit lifecycle she built — and how it's still evolvingWhy breaking work into focused, separate chats is the key to quality outputHow human expertise remains non-negotiable in the review processWhat it feels like to manage a "team" of AI agents instead of humansOh, and baby Willow made a cameo. That's just how we do things here.Find Clea on LinkedIn to keep the conversation going: https://www.linkedin.com/in/cleamahoney/ Connect with Samantha on LinkedIn:https://www.linkedin.com/in/samantha-murray613/   Chapters:00:00 Introduction03:44 Clea's Journey in Customer Education06:34 Understanding Roost and Its Challenges08:50 Adapting to Team Changes and AI Integration11:04 Transitioning to Claude AI for Workflow Design13:40 Developing Effective Curriculum with AI18:47 Iterating on Course Design and Content Creation21:59 Incorporating Feedback for Quality Content24:38 Adapting Terminology for User Understanding26:49 Breaking Down Complex Processes30:41 Automating Course Updates and Maintenance35:48 The Role of Human Expertise in AI37:53 Managing AI Agents as Team Members41:37 Continuous Optimization in Learning Design