RAG-A

• GIF

Vision

Our project is driven by the vision of empowering individuals to make informed and healthy food choices effortlessly.

Functionality

Our application simplifies the process of decoding nutritional information found on packaged food labels, aiding users in making well-informed decisions about their food consumption.

Development Approach

Utilizing the Gemini-Pro-Vision model, we've crafted a solution that can analyze images uploaded by users, extracting pertinent data from food labels. Complementing this, we've implemented a RAG-based system to cross-reference the extracted information with recommended dietary allowances and other crucial data, delivering insights into the nutritional content of the food.

Overcoming Challenges

Navigating through engineering challenges, we fine-tuned prompts to accurately identify food labels and optimized image preprocessing for seamless integration with the Gemini models.

Achievements

We take pride in the tangible impact of our application, empowering users to adopt healthier dietary habits through easy access to nutritional information.

Key Takeaways

Our journey taught us invaluable lessons in teamwork, adaptive learning, and the intricacies of developing user-centric GenAI applications from scratch.

Future Prospects for RAG-A

Looking ahead, we aim to enhance the performance of our application and extend its reach by developing a mobile version, ensuring accessibility and convenience for a wider audience.

Built With

• fastapi
• langchain
• python
• unstructured.io