Customer story · Code Girls United
Student-made AI lessons, ready for any classroom.
How Code Girls United uses Claude to turn its library of student-created AI Academy videos into classroom-ready lesson plans and a teacher guidebook, with a curriculum workflow built by AIFluency.ai.
Education & Learning · Implementation · Process redesign
The organization: Code Girls United
Hands-on technology, coding, AI, cybersecurity, and business education for girls across Montana, including rural and Tribal communities.
- Industry: Nonprofit education
- Business unit: Program Delivery & Curriculum
- Technology: Claude Team · Claude Project · Skill
- Delivered by: AIFluency.ai · Led by Cody Rose
Visit Code Girls United
- 32 hrs of staff time to build each lesson
- 30+ lessons waiting in the backlog
- 900+ staff hours the backlog represented
Their story: It started with 28 girls in a diner.
Code Girls United was founded in 2016 after its founder saw firsthand how few girls were represented in college computer science classrooms. In the fall of that year, 28 girls brought their laptops to the basement of Sykes Diner in Kalispell to learn to code.
That gathering has grown into a statewide nonprofit. Girls get hands-on education in technology, coding, AI, cybersecurity, and business that builds confidence, creativity, problem-solving skills, and pathways to future careers.
Its AI Academy is one of the fastest-growing parts of that work. The girls create the lesson videos themselves.
01 / The challenge: The curriculum existed. The hours did not.
The AI Academy grew quickly, and so did its library of student-created videos. Turning each video into a classroom-ready lesson took a staff member roughly 32 hours: reviewing the recording, organizing the content, writing activities and discussion questions, and building the lesson plan and teacher resource.
More than 30 lessons were waiting. That added up to 900+ hours of staff time, in an organization already running programs across Montana. As finished lesson plans, the same material could reach educators and students well beyond the girls in Code Girls United’s own programs.
AIFluency.ai’s role was to build a repeatable curriculum-production workflow, then train the person who would run it. The starting point was material Code Girls United already owned.
More than 900 hours of lesson development stood between a growing video library and the classrooms that could use it.
02 / What we built: One workflow, built inside Code Girls United’s own workspace.
AIFluency.ai configured a Claude Project and built a Skill inside Code Girls United’s Claude Team workspace. Together, they carry the organization’s curriculum context into a consistent lesson-plan format.
A workspace the team controls
Claude Team provides the working environment under Code Girls United’s account, with access for the staff who run the workflow.
Context that stays with the work
The Claude Project holds curriculum guidance, organizational voice, and the appropriate level for the intended age group.
A consistent lesson-plan format
The Skill turns a video transcript and its associated activity into a structured lesson plan and teacher resource, ready for staff review.
Review, training, and handoff
The first lessons went through staff review and revision. AIFluency.ai delivered written operating instructions and trained the staff operator in a live session.
03 / How it works: From a student-made video to a lesson any educator can teach.
Staff supply the transcript and activity. Claude produces a draft in the agreed structure. Code Girls United reviews and validates every lesson before it reaches a classroom.
- Source. The transcript of a student-created AI Academy video, plus its associated activity.
- Draft. The Claude Project and Skill apply curriculum context, age-level guidance, and the lesson-plan structure.
- Review. A staff member checks accuracy, clarity, age appropriateness, activities, and answers, then revises and approves the lesson.
- Release. Approved lessons join the lesson-plan series and teacher guidebook for educators.
Each lesson includes: learning objectives, concept explanation, worked example, activity, discussion questions, understanding checks, answer key, teacher notes.
04 / The result: Staff improve lessons instead of starting from scratch.
Code Girls United now has a repeatable way to work through its backlog. Each lesson starts from a Claude draft, and staff spend their time improving and validating the material. Every lesson is reviewed and approved before it reaches a classroom.
Preparation time: A first draft in minutes, where each lesson used to take about 32 hours. Staff time now goes to review and validation instead of building every lesson by hand.
How the curriculum workflow changes| Area | Before | With the workflow |
|---|
| Preparation | About 32 staff hours to build each lesson by hand. | Claude drafts the lesson plan and teacher resource for staff review. |
| Backlog | 30+ lessons and 900+ hours of work waiting. | A working process for clearing the remaining lessons. |
| Operation | Progress depends on finding time for each lesson. | A trained staff operator runs the workflow directly. |
| Quality | Staff develop and check the material. | Staff review and validate every lesson. |
In her words: Marianne Smith, Founder & CEO, Code Girls United
“Code Girls United’s AI Academy has grown quickly, but turning engaging student-created videos into consistent, classroom-ready lessons has been a challenge. This project uses Claude to help transform our growing library of AI Academy videos into a practical lesson-plan series and teacher guidebook, making the content easier for educators to use and expand our reach beyond the girls who participate directly in our programs.
“For the girls, it also demonstrates something powerful: AI is not just a tool they consume, it is a technology they can use to create, solve problems, and build resources that help other students learn.”
About Marianne Smith
Marianne Smith is the founder and CEO of Code Girls United (CGU), a Montana-based nonprofit redefining access to computer science, artificial intelligence, and business education for girls in rural and Tribal communities.
An engineer with BS and MS degrees, Marianne brings more than 35 years of experience across technology, business, and education, including work as a NASA engineer and as an Adjunct Professor of Computer Science at Flathead Valley Community College. She is the architect of CGU’s AI Academy, a rigorous, project-based program that integrates AI literacy, ethical reasoning, real-world problem solving, and entrepreneurial thinking. Students learn to build technology and to understand how it becomes products, companies, and solutions in the real world.
A recognized leader in AI and technology education, Marianne is committed to advancing responsible, human-centered AI and preparing girls to become the engineers, innovators, founders, and leaders who will define the future.
05 / Training & ownership: Built for the people who will use it.
Delivery included review rounds on the first lessons, a one-page operating guide, and live training for the staff operator. Code Girls United now runs the process in its own workspace.
The organization owns its source curriculum and every lesson produced, and it keeps using the Project and Skill without relying on AIFluency.ai for each new lesson.
The workflow handles curriculum only. No student data goes into it.
06 / About the architect: The architect behind the workflow.
Ryan Woodard, PhD
AI & machine learning engineer · Bozeman, Montana · Claude Certified Architect, Foundations
Ryan architected the curriculum workflow for this AIFluency.ai engagement. A physicist and machine learning engineer, he brings six years of quantitative research at ETH Zürich and a decade of experience putting machine learning into production across fraud detection, language processing, and computer vision.
His recent work spans the design, training, evaluation, and deployment of generative models. For Code Girls United, he applied that experience to a practical goal: a repeatable process staff run in their own Claude workspace.
Architect: Ryan Woodard. Delivery lead: Cody Rose, AIFluency.ai
What is waiting on your team?
Tell us where the work gets stuck. We’ll help you identify a practical place to start. Talk with AIFluency.ai
