A faith & fitness app taken from concept to a fully deployed, database-backed product, designed and built by one person, using AI at every stage.
I started with competitor research across the faith & fitness space, then used ChatGPT for initial concepts and branding, landing on a name that felt modern and not preachy. I validated directions with fitness-minded and spiritually curious users, which shaped the name and the two-tone FITFUL / MIND wordmark.
Starting with the Bible app and top fitness apps during the discovery phase.



Early concepts used the lion and the working name GodlyFitFam to represent faith, strength, and community. Based on user feedback, I simplified the identity and shifted to the crown as a symbol of purpose, inner strength, and achievement. The result felt more modern, versatile, and better suited for the product.
→
→
→
→
The core user skews female, so directionally I’m starting to roadmap more features that cater to the primary users, then widening the brand through tone and theme options to welcome the secondary audience.

Trains 4 to 5 times a week · prays daily · reads scripture and self help books most days
What version of the Bible are you using?
Why she is the core user: women are about 60% of fitness app users, buy about 74% of self help books, and are likelier to pray daily (50% versus 37% of men).

Gym warrior · leans on spirituality to stay grounded · the gym is his safe space
I'm not really religious, so the old name was a turn-off.
Why he matters: 78% of people who exercise train mainly for mental well-being, so his way in is the “Mind,” not the faith.
I started with ChatGPT to explore the concept, define the core experience, and work through early product thinking. From there I used Figma Make to build out the interface, refine the visual system, and work through key user flows. That gave me a polished prototype I could evaluate as a real experience rather than a collection of static screens.
Once the foundation was in place, I moved into Claude Code where I could push the product further. Instead of treating AI as a one-time generator, I used it as part of an ongoing design and development loop. I could build, review what was actually working, make decisions, and refine the experience in real time.
Once the concept proved itself, I moved beyond prototyping and built FitfulMind as a working product with a production server and MySQL database, a GitHub workflow, and an SSH based deployment process. The result was more than an AI generated prototype. It became an exercise in how AI can connect product thinking, design, development, and deployment into one continuous workflow.




Streaks & badges. Daily visit tracking and persistent Bronze, Silver, Gold, and Diamond badges that survive streak resets. God Mode. A daily challenge system that unlocks when users complete three or more disciplines.
Saved Scriptures. Users can bookmark meaningful scriptures and build a personal collection to revisit for encouragement, reflection, or inspiration. Daily Notes. A simple space to capture thoughts, prayers, gratitude, and reflections alongside their daily faith and fitness routine.


Commentary Tone plus Light & Dark themes. Choose how daily insights are delivered, from Cool Spiritual for calm reflection to Fired-Up Coach. Switch between light and dark modes to create a visual experience that fits your preference.
The live web app, ready on any phone or desktop browser.
The public site introduces FitfulMind to people who have never heard of it. It frames the product around training the mind and building the body, explains the daily scripture and the two coaching tones, and moves visitors into a quick onboarding and straight to their first daily verse.
The live landing page, built to educate first-time visitors and turn them into users.
Behind the app sits an admin dashboard for managing users, editing scripture and commentary, reviewing feedback, watching analytics, and testing badges and streaks. It is how one person operates everything.
Password-protected interactive demo with sample users and feedback. Ask me for access.
Using Claude connected to Figma, I built a full design system that references the production codebase, with tokens, components, and interaction states that reflect what's actually live, not a theoretical spec. Then I used the Figma Agent to clean it up and align it to production.
Typography & Color System
Primary typeface for system and UI. Weights: Light · Regular · Medium · Semibold · Bold. A B C D E F G H I J K L M N O P Q R S T U V W X Y Z 1 2 3 4 5 6 7 8 9 0
A high-impact condensed sans-serif used for bold marketing statements, hero banners, and section highlights. It delivers editorial presence and strong hierarchy.
Chosen to balance the condensed Bebas Neue. Warm, geometric, and highly readable, it governs brand storytelling, body copy, and secondary callouts.
The workhorse system font for the product UI and operational interfaces. Pixel-perfect readability, dense layouts, and clean structure across components.
With a real product live, testing kept going. Feedback from gym-goers and spiritually minded users continued to shape the tone options, the scripture translations on the roadmap, and the themeable UI. It is the same loop that first drove the rebrand from GodlyFitFam to FitfulMind.
I'm not really a religious person, so “GodlyFitFam” is kind of a turn-off.
◆ Drove a changeWhat version(s) of the Bible are you using?
◆ Drove a changeNice and clean and easy to understand.
✓ ValidatedNice that there are options for the tones of the motivation coach commentary.
✓ ValidatedThe next step is packaging FitfulMind as native iOS and Android apps and launching on the App Store and Google Play. Beyond the app, there are opportunities to extend the brand through community, content, apparel, and real-world experiences. These early concepts explore what that broader activation could become.
Working across Claude Chat and Claude Code, I learned to run AI like a real project. Every change was versioned in build snapshots, one staging folder stayed the single source of truth, and I always had a history to roll back to. Markdown handoff files kept context and decisions in writing, which cut down on hallucinations. Starting fresh sessions from those files kept the work fast and saved tokens.
Knowing when the output was right, when it needed redirecting, and when to override it entirely. The prompting craft is itself a design skill.
Work that would take weeks or months took days. When the cost of a change drops, you test more ideas, kill bad ones faster, and arrive at better solutions than a fixed-scope build allows.