'Roo preview

'Roo

A workout tracking app I built for myself and my family. I wanted a way to track our fitness progress without handing over personal health data to big tech companies. Built with Next.js and React Server Components for fast, server-rendered pages with minimal client-side JavaScript. The UI uses shadcn/ui and TailwindCSS for a clean, accessible interface. For the database, I chose Turso—a SQLite-based edge database that keeps costs low while providing excellent performance. The whole thing is deployed on Vercel. The source code is fully open source, so anyone can self-host their own instance.

Technologies

Next.jsTailwindCSSshadcn/uiTursoReact Server ComponentsVercel

AI Program Generator

I built this app for my family, who aren't familiar with the intricacies of fitness programming. To help them get started, I created an AI Program Generator that builds personalized workout plans tailored to their goals, schedule, and preferences. Under the hood, I integrated GPT-4o-mini via the OpenAI SDK to analyze inputs like fitness objectives, available training days, and specific needs—then generate a complete, structured program ready to follow.

Your Custom Program

Here's what a generated program looks like. The AI produces a structured plan broken down by day, with specific exercises, sets, reps, and form cues for each movement. Users can regenerate if they want something different, or save it directly to their programs list and start training right away.

Weekly Analytics

Keep tabs on your training with a simple weekly breakdown. See which days you trained, how many sets you completed, and track your consistency over time. Nothing fancy—just the data you need to stay accountable and see your progress at a glance.

Nutrient Tracker

Log your daily nutrition alongside your workouts. Track calories, protein, carbs, and fats to make sure your diet supports your training goals. The interface keeps it simple—just punch in your meals and watch the totals add up throughout the day.