Good Inside · 2025–Present

Transforming Long-Form Content into Micro-Learning Journeys

The challenge - App completion sat below 20%, and parents couldn’t tell where the journey was taking them.

The innovation - AI agents - Model Context Protocol plus a Collection Builder - refactor a deep content library into modular, personalized “Collections.”

The impact - A dozen-plus Collections at 7X speed with 2× completion, later spun into a standalone product.

Deeply Feeling Kids product page with Dr. Becky video

The Challenge

When I joined Good Inside, completion of the app journey’s “focus areas” sat consistently below 20%, and qualitative feedback showed parents were unclear “where the app was taking them.”

I determined the fix wasn’t net-new content - it was refactoring the existing library into modular learning “Collections” that assemble into a personalized journey around the topics each parent cares about most.

The Approach

I rebuilt the content system from the ground up, leading design and build across all three phases of a new parent journey: a new in-house content management system, the front-end experience, and content production.

The learning-design solve was to refactor the existing app content - recorded Workshops, parenting scripts, and interactive decks - into “Collections” that each drive toward a clear parenting outcome (e.g. “What to do about your kid’s resistance to getting out the door in the morning”).

Good Inside content framework - how a concept shows up in your home — Every Collection drives toward a clear, in-the-home parenting outcome.

The Build

AI played a key role in every phase. Using Model Context Protocol, an agent automated large portions of the content migration.

I’d input a parent’s need and the Collection Builder agent did the rest - pulling short clips from long-form Workshops, finding relevant scripts, assembling interactive decks, and coordinating with the internal chatbot to keep every title, description, and line true to the Good Inside voice.

Once I reviewed and approved the content in Claude CoWork, a third agent wrote it - again via MCP - straight into the content management system, ready to push to the app.

A Collection: “When Resistance Shows Up in the Morning”

Program view - sectioned short videos and a script

A story-based scenario card: “Morning Refusal”

Parent script interstitial: talking to your DFK after a meltdown

The Collection Builder’s output: assembled Collections, sectioned shorts, story-based cards, and on-brand parent scripts.

The Impact

As a team of one plus AI, I delivered the work of an entire team - over a dozen Collections at 7X the speed of the previous manual workflows.

The new Collections more than doubled completion rates, and engagement data showed parents didn’t just finish one - they moved on to multiple Collections. And because the Collections are modular, our AI can now curate personalized journeys from the content.

The Collections were so successful that I led the development and launch of standalone Collections sold as their own product - a lower-commitment entry point into the subscription membership.

← PreviousExpert Knowledge into Elevated Learning Experiences Next →Soft Skills “Accelerator” for Frontline Workers