HealthMate

Your AI Coach for a Healthier, Smarter You.

Comment

🧠 About the Project — FitMate AI

"Your Lifestyle. Your Plan. Powered by AI."

💡 Inspiration

In today’s world, fitness and wellness have become more of a trend than a personal journey. But every person’s body, routine, and mindset are different — so why should their fitness plan be the same? This question inspired us to build FitMate AI, an app that uses Artificial Intelligence to understand you first, and then create a fully personalized plan for your gym, diet, sports, or daily wellness routine.

⚙️ How We Built It

We designed a simple yet intelligent flow:

User Selection: The user starts by selecting a purpose — Gym 🏋️, Diet 🍎, Sports ⚽, Mental Health 🧘, or General Wellness 🌿.

Goal Input: The user writes their goal manually, e.g.,

“I want to lose fat and gain muscle within 3 months.”

AI Analysis (Gemini): The goal is processed by Google Gemini API, which generates a set of 20 multiple-choice questions (MCQs) to deeply understand user behavior, lifestyle, diet, and habits.

User Interaction: The user answers these MCQs in a gamified interface.

AI Plan Generation: Based on the responses, Gemini analyzes the data and generates a complete personalized plan including:

🏋️ Workout schedule

🍽️ Diet chart

⏰ Daily routine

💬 Motivation and wellness recommendations

Plan Delivery: The user can download the plan as a PDF or share it on social media or messaging apps.

🧩 Tech Stack Layer Tools / Technologies Frontend React / Next.js + TailwindCSS + shadcn/ui Backend Node.js (Express) Database PostgreSQL / MongoDB AI Layer Gemini API (LLM for Q&A + Plan Generation) Auth Firebase Authentication PDF Generation pdfmake / Puppeteer Storage Firebase / Cloudinary Hosting Vercel + Render 🚧 Challenges We Faced

🧠 Designing prompts that make Gemini ask smart, context-aware questions.

⚖️ Balancing between user freedom (manual goals) and structured AI flow.

🎨 Making the UI simple yet motivating — no “boring fitness form” vibe.

💾 Managing user data securely while generating and storing personalized plans.

🕒 Handling API latency and ensuring real-time AI responses without lag.

🎓 What We Learned

How to create interactive AI pipelines combining LLMs and user input dynamically.

The importance of behavioral context in fitness recommendation systems.

Prompt-engineering techniques to get accurate, goal-specific responses from Gemini.

How AI personalization increases engagement and motivation for health users.

🧮 Core Logic (Simplified Formula)

The AI computes a Personalization Score (P) for each user based on their answers:

The AI then uses this score to fine-tune diet, workout, and lifestyle recommendations.

Share this project:

Updates