Day with AI: Fitness App, Security Course, and a Static Site — Three Projects in One Day.
Complete fitness app rebuild, interactive security course with quizzes, and a legacy site fix. 4 commits, 6 sessions.

What I worked on
Thursday. One of those days where everything lands at once — three completely different projects from morning to night.
What I did
Fitness app — complete rebuildBiggest chunk of the day. Took an existing fitness app and rebuilt it from scratch. New design system, four-day training plan, AI coach, meal plan with supplements, countdown timer, and PWA support. From "it kind of works" to "I'd actually open this every day." Without AI? At least a week. With AI? Four hours.
Interactive security courseComplete web application with 12 chapters on online security. Each chapter has a quiz so you can verify understanding. Passwords, phishing, two-factor auth, social engineering — all explained clearly with interactive testing. From zero to deployed in about an hour.
Legacy site — static build and deployAn older web project that needed its hero section photos fixed and a case-sensitivity bug resolved. Classic problem — works locally, breaks on the server. Plus setting up the deploy script.
AI images and videos — prompt tuningSession focused on getting better results from generative models. Rules for composition, lighting, aspect ratios, and expression stability for follow-up animation. Practical know-how useful for the entire content pipeline.
Cold email pipelineMorning automated run. Scraping, data enrichment, draft generation. Runs in the background.
Time: AI vs without AI
| Task | Without AI | With AI |
|---|
| Fitness app — complete rebuild | 8+ days | ~4h |
|---|---|---|
| Security course (12 chapters + quizzes) | 8+ days | ~1h |
| Legacy site — fix + deploy | 3-4h | ~2h |
| AI prompts — images and videos | 2-3h | ~1h |
| Cold email pipeline | — | ~10min |
| Total | ~20 days | ~8h |
What I learned
- Three completely different projects in a day isn't a problem when you have AI as a copilot — context switching hurts much less
- Case-sensitivity bugs are sneaky — works on Mac, breaks on Linux. Always test on the target environment
- Interactive courses with quizzes get much higher engagement than static text — and with AI you create them in a fraction of the time
- A fitness app with an AI coach is an interesting use case — training personalization is exactly the kind of task where AI excels
Interested in the article?
Let's discuss what this kind of automation can do in your company.
Free consultation