AI-Powered Workout Tracker
This is a comprehensive full-stack gym and workout tracking application designed with a mobile-first PWA approach. The frontend is powered by Next.js, React, and Tailwind CSS, while the backend relies on Django REST Framework. The system allows users to seamlessly register, create customized workout plans, and track live sessions including sets, reps, and weights. A standout feature is the AI and OCR integration, enabling users to generate workout routines from text prompts or extract them from uploaded documents. The robust architecture distinctly separates reusable workout templates from historical performance data, providing powerful analytics on volume trends, estimated 1RM, and overall adherence.
Year
2026
Role
Full Stack Developer
Technologies
Next.js, React, Django REST Framework, Python, TypeScript, Tailwind CSS, OpenAI API, OCR, Docker, PWA
Challenge
The main challenges included designing a nested data model to separate workout templates from live session history without mixing planned targets with completed performance. Additionally, managing secure session-based CSRF authentication across a separate Next.js frontend and Django backend, handling real-time data updates during active workouts, and normalizing unpredictable AI/LLM outputs into a stable, frontend-ready JSON schema were significant hurdles.
Solution
Developed a normalized backend schema and REST ViewSets to handle nested templates and performance records separately. Implemented a robust session/CSRF authentication flow between the frontend and backend APIs. Isolated the AI workout generation into a dedicated backend module with strict validation and normalization before database insertion. Finally, the entire application was containerized using Docker and docker-compose for seamless local development and deployment.
Results
Delivered a complete, mobile-first PWA fitness tracker with secure authentication.
Integrated AI and OCR workflows for automated workout plan generation and import.
Built advanced analytics to track volume changes, progress trends, and plan adherence.
Engineered a scalable backend that cleanly separates workout templates from performance logs.
Fully containerized deployment using Docker.