Meal Planning Made Simple.

Discover recipes customized to your dietary needs with NutriGenius AI — your smart companion for finding the perfect meal.

🏛️System Architecture

NutriGenius AI uses a microservices architecture for scalability and maintainability. Hover over a service in the diagram below to learn more about its specific role and technology.

👤 Frontend (Next.js)
🛡️ User Authentication
🚪 API Gateway (FastAPI)
👤 User Profile
👨‍🍳 Recipe Data
🧠 ML Inference
💾 Database (Appwrite)

Hover a Service

Detailed information about the selected service will appear here.

🗺️The Data Journey

Follow the path of data from user onboarding to receiving a personalized meal plan.

1

Onboarding & Data Collection

The user registers and provides profile data (biometrics, goals, medical history) via a guided wizard. Data quality is ensured with client and server-side validation.

2

Secure Storage

The user's data is securely stored in the database, which is used to train the ML models. The data is encrypted at rest, ensuring privacy and security.

3

Recommendation Request

The user fills out a form to request a personalized meal plan, including dietary preferences, health goals, allergies, and other factors.

4

AI Generation

The ML service fetches the user's profile and interaction history, feeding it into the hybrid model to generate a ranked list of suitable recipe IDs.

5

Displaying Results

Recipe IDs are "hydrated" with full details (name, image, ingredients). This rich data is sent to the frontend and rendered in an interactive dashboard. The user can then select and view each recipe individually.

🧠The AI Core

Our system fuses multiple high-quality datasets into a single, optimized source of truth. This data-driven foundation allows our intelligence engine to deliver fast and deeply personalized recommendations.

A Fused, High-Quality Master Dataset

We combine data from multiple sources—including our core Indian Recipe Dataset and a dedicated Indian Nutritional Database—into a single, pre-compiled master file.

Key Attributes Analyzed:

📖RecipeName
🌿Cleaned_Ingredients
🌶️Cuisine
❤️Diet
🔥Calories (kcal)
💪Protein (g)
rating_avg
📊n_ratings

⚙️How It Works

A look under the hood at the core algorithms and strategies that power our AI, ensuring fast, accurate, and relevant recommendations.

Text Analysis & Retrieval

The initial search uses a classic NLP combination of TF-IDF to identify the most important ingredients in each recipe, and Cosine Similarity to find the recipes that best match the user's text query. This is highly efficient for keyword-based retrieval.

Data Merging & Enrichment

To combine our recipe data with the Indian nutritional dataset, we use Fuzzy String Matching with the RapidFuzz library. This allows the model to intelligently match recipe names (e.g., 'Paneer Butter Masala') to their nutritional profiles, even if the names aren't identical.

Rule-Based Filtering

Before ranking, a deterministic filtering layer removes any recipes that don't match the user's hard constraints, such as Cuisine Preference, Dietary Needs, or maximum Cooking Time. This ensures all results are relevant.

Personalized Ranking

The core of the system is a Weighted Multi-Factor Scoring function. This heuristic-based Learning to Rank (LTR) approach combines multiple features—text similarity, nutritional alignment with user's BMI and goals, and course suitability—into a single score to find the truly best recommendations.

Advanced Features

Cutting-edge integrations transform the application from a simple tool into a dynamic and intelligent cooking companion.

🧠

Dynamic Nutritional Analysis

Leverages a focused dataset of Indian cuisine to enrich every recipe with accurate nutritional data, including calories, protein, and fiber. This allows for true health-based recommendations.

🎯

Hyper-Personalized Ranking

Goes beyond keywords to rank results based on a multi-factor score, considering your BMI, health goals, cooking skill, and even ingredients you already have in your pantry.

🍲

Context-Aware Recommendations

The model understands the difference between a snack and a main course. When you ask for dinner, it intelligently prioritizes proper meals over appetizers to give you more suitable results.

⚡️

Optimized for Speed

All heavy data processing and merging is done once at startup. This creates a single, efficient master dataset, ensuring your recommendations are generated almost instantly.

Meet the Team

The dedicated individuals bringing NutriGenius AI to life.

Vedant Bhor

Vedant Bhor

Project & Full-Stack Lead

    - Next.js Full Stack Development
    - Training & Tuning ML Models
    - Model Performance Optimization
    - Implementing Core Algorithms
GitHubLinkedInDiscordPortfolio
Tanay Hingane

Tanay Hingane

AI & Authentication Specialist

    - Appwrite Database Management
    - User Authentication with Clerk
    - Meal History Feature Logic
    - User Profile Management
GitHubLinkedInDiscordPortfolio
Adarsh Tile

Adarsh Tile

Backend & API Lead

    - Developing FastAPI Backend
    - ML Model & API Integration
    - API Endpoint Management
GitHubLinkedInDiscord
Sushrut Deshmukh

Sushrut Deshmukh

Core Algorithm Lead

    - Suggesting Core Algorithms
    - Model Performance Optimization
    - Project Report Writing
GitHubLinkedInDiscord
Yadnesh Udar

Yadnesh Udar

QA & Testing Lead

    - End-to-End Project Testing
    - Unit & Integration Test Cases
    - Overall Quality Assurance
GitHubLinkedInDiscord

📚Citations

This work is inspired by and builds upon recent advancements in personalized nutrition and intelligent health systems.