Menu-Sahayak：印度餐廳餐牌 AI 過敏原檢查工具

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle.

What I Built

Menu-Sahayak is an open-source AI-powered allergen checker designed for Indian restaurant menus.

Eating out in an unfamiliar place is challenging for people with food allergies. The problem becomes harder for travelers who struggle to understand Hindi or regional-language dish names and their ingredients.

A translation tool might explain what Chole Bhature is, but it does not necessarily identify its potential allergen risks.

Menu-Sahayak helps bridge this gap.

Users paste restaurant menu items, one dish per line, and the app returns a numbered list containing:

Dish names.
Likely allergens, including dairy, nuts, gluten, eggs, and soy.
A concise safety note describing potential allergen sources.

Powered by Llama 3.3 70B through Groq, the app analyzes Indian dishes and identifies potential allergens based on common ingredients and preparation methods.

My goal is to help travelers, local diners, and people with dietary restrictions make more informed food choices.

Menu-Sahayak provides preliminary allergen guidance, not a guarantee of food safety. Ingredients, recipes, and cross-contamination risks vary between restaurants. People with severe allergies should always verify ingredients and preparation practices directly with restaurant staff.

Demo

The interface allows users to enter multiple Indian dishes and review their potential allergens and safety notes in a structured format.

Project Repository: Menu-Sahayak on GitHub

Code

Allergen Lens — Menu Allergen Checker for Indian Restaurants

A lightweight Gradio app that analyzes Indian restaurant menu items and flags likely allergens per dish.

What it does

Paste menu items (one dish per line), and the app returns a numbered list with:

dish name
likely allergens from: dairy, nuts, gluten, eggs, soy
one concise safety note

The model is instructed to avoid unsupported certainty: allergens are marked as likely/possible based on common preparation, and users should verify with restaurant staff for severe allergies.

Why this is useful

Tourists with allergies may struggle to read Hindi or regional-language menus. This tool provides quick, practical allergen guidance dish-by-dish.

Setup

Clone this repository.

Install dependencies:

pip install -r menu-sahayak/requirements.txt

Create env file:

cp menu-sahayak/.env.example menu-sahayak/.env

Add your Groq API key in menu-sahayak/.env:
```
GROQ_API_KEY=your_key_here
```
Run:
```
cd menu-sahayak
python app.py
```

Get a free Groq…

How I Built It

I built Menu-Sahayak using Python, Gradio, and Llama 3.3 70B, accessed through Groq's inference API.

AI Model: Llama 3.3 70B provides the language understanding used to analyze dish names, common ingredients, and potential allergens.

Frontend: Gradio provides a lightweight web interface for entering menu items and displaying the results.

Core Logic: The application sends menu items to the language model with a focused system prompt that instructs it to act as an allergen analyst. It produces numbered dish-level results with likely allergen labels and concise safety notes rather than unrestricted conversational responses.

Error Handling: The application includes handling for empty inputs, missing API keys, and API errors or rate limits.

Configuration: Environment variables keep the Groq API key separate from the application code. Developers configure their own key through a local .env file.

I kept the implementation lightweight so developers can set it up, test it, and extend it without needing a complex infrastructure.

Tech Stack

Python for application logic.
Gradio for the web interface.
Llama 3.3 70B for AI-powered allergen analysis.
Groq for hosted model inference.
Environment variables for API configuration.

Why Does Open Innovation Matter?

Open innovation makes it easier for independent developers to experiment with AI and build practical tools around real problems.

1. Access to Open-Weight AI

Llama 3.3 70B provides the language capabilities needed to analyze Indian dish names and common ingredients. Its open-weight availability, subject to its license, gives developers more flexibility over how they evaluate and integrate the model.

2. Lower Barriers to Experimentation

Groq provides hosted inference with pricing and usage limits determined by its current service terms. This makes it easier to prototype an AI-powered side project without immediately managing GPU infrastructure.

3. Flexibility and Model Choice

Menu-Sahayak keeps its analysis instructions separate from its user interface. This provides a starting point for evaluating other compatible models and inference providers, with code or prompt adjustments where necessary.

4. Transparency and Community Contributions

The public source code allows developers to inspect the implementation, report issues, and contribute improvements. Future extensions include regional-language support, better ingredient explanations, and testing across a wider range of Indian cuisines.

For a tool related to dietary safety, open development also provides an opportunity to review limitations, test outputs, and improve how uncertainty is communicated.

What's Next?

I plan to explore:

Support for Hindi and regional-language menus.
More detailed ingredient explanations for Indian dishes.
Improved testing across regional cuisines and variations in recipes.
A simpler deployment process so more people can try the application.
Community contributions to improve usability and allergen guidance.

My aim is to make Menu-Sahayak a practical, accessible tool for people who need help understanding potential allergens in Indian restaurant food.

Prize Categories

Overall Side Hustle Challenge Entry.