Home Analyzer About

👋 Meet the Team

Passionate builders using AI to make nutrition accessible for everyone.

H

Harika

Co-Founder & AI Engineer

Specialises in computer vision, model fine-tuning, and deep learning. Trained the custom 80-class Indian food model and built the 3-model ensemble pipeline.

🤖 ML & Deep Learning 👁️ Computer Vision 🐍 Python & PyTorch 🎓 Model Fine-Tuning
D

Deepika

Co-Founder & Full-Stack Developer

Expert in Flask, REST APIs, and modern frontend design. Built the entire NutriVision web application, UI/UX, and API integrations.

💻 Full-Stack Dev 🎨 UI/UX Design ⚡ Flask & REST APIs 🌐 JavaScript
N

Nageswari

Co-Founder & AI Engineer

Focused on AI-driven nutrition analysis and intelligent recommendation systems. Passionate about using machine learning to improve health awareness and decision-making.

🧠 Deep Learning 📊 Data Analysis 🤖 AI Systems 🥗 Nutrition AI
Our Purpose

The Mission 🎯

We believe that understanding what you eat should be simple, instant, and free. NutriVision combines state-of-the-art computer vision with the world's best open-source LLMs to make nutrition analysis as easy as taking a photo. Our focus on Indian food — a severely underrepresented domain in food AI — is what sets us apart.

Under the Hood

Technology Stack 🛠️

👁️

nateraw / food

Food-101 dataset — 101 international food categories, primary classifier.

🍛

Indian-Western-34

34 specialised Indian & Western food classes — second classifier layer.

🎓

Custom 80-Class Model

Fine-tuned in-house on rare Indian regional dishes. Covers Adhirasam, Ghevar, Ariselu & 77 more.

🤖

Llama 3.1 8B

Via OpenRouter — fast, accurate nutrition data and personalised health insights.

Flask + GPU

Python Flask backend running on CUDA T4 GPU for fast vision inference.

🎨

Modern Frontend

Responsive, animated UI with CSS animations, Syne font, and dark theme.

Differentiators

What Makes Us Different ✨

01

Indian Food Focus

Covers 80+ regional Indian specialties that most food AI completely misses — from Rasmalai to Pootharekulu.

02

3-Model Ensemble

Running 3 independent classifiers simultaneously and picking the best confidence result — not just a single model.

03

Personalised Reports

Every report is tailored to your age, gender, BMI, dietary preference, and health condition.

04

No GPU Overload

Vision models run on GPU; LLM is handled by OpenRouter cloud — balanced resource usage on T4.

Ready to Try NutriVision? 🚀

Experience the future of AI food analysis — free, fast, personalised.

Start Analyzing Now →