Swasth - The Fitness Companion

                                                      Swasth – The AI Fitness Companion

In today’s fast-paced world, maintaining fitness can be difficult. As college students, we saw many people around us struggling to stay healthy because of busy schedules, lack of motivation, and no proper guidance. Even we faced the same issue during our studies. That’s what inspired us to create Swasth, an AI-powered fitness web app that makes workouts fun, interactive, and easy to follow.

Swasth helps users stay active by turning ordinary workouts into enjoyable, AI-driven experiences. Using your webcam, the app can detect exercises like squats, planks, and jumps, count repetitions, and even provide real-time feedback on your form. It also tracks calories burned and displays your progress instantly. The platform includes fitness games such as Squat Race, Plank Duel, and Jump Wars, where users can challenge their friends or compete globally. This makes exercising more engaging and helps build consistency through friendly competition.

The app features a personalized dashboard that shows daily, weekly, and monthly activity progress. Users can also interact with an AI fitness coach that suggests meals, workout tips, and motivation throughout their journey. To keep the community connected, Swasth offers two leaderboards — one for friends and one global — where users can track their ranks, celebrate achievements, and push each other toward their goals.

We built Swasth using React and TailwindCSS for the frontend and Node.js with MongoDB for the backend. For AI pose tracking, we used Google AI Studio, TensorFlow.js, and MediaPipe, which made implementing movement detection much easier. Google AI Studio helped us integrate machine learning models efficiently without needing complex manual setup. We also used Socket.IO to enable multiplayer modes and real-time leaderboard updates.

While building Swasth, we faced some challenges, but tools like Google AI Studio really simplified the process. It helped us handle pose tracking and model integration with minimal configuration. Our main challenge was fine-tuning the accuracy of movement detection and ensuring smooth performance during multiplayer sessions. Once we optimized these parts, the app started running much better.

Creating Swasth taught us how to combine AI with web development effectively. We learned about pose estimation, real-time systems, and user engagement through gamification. More importantly, we discovered how technology can inspire people to stay consistent and make fitness something to look forward to every day.

Looking ahead, we plan to add AI-personalized workout routines, voice-guided coaching, and integration with wearable devices for detailed analytics. We also want to launch a mobile version and bring in team-based challenges so users can work out together.

In short, Swasth isn’t just a fitness app — it’s a digital fitness companion that combines AI, motivation, and community to help people live healthier and more active lives. Swasth = Fitness + Fun + AI + Motivation