Inspiration
Our Nutri AI is here to reshape the way we approach food. In a world dominated by fast and ready-made meals, our bot becomes your personalized meal planner and recipe generator right on your phone. Whether you're a pregnant woman seeking nutritious options or a budget-conscious gym enthusiast craving diverse and protein-rich dishes, our bot is your all-in-one solution for a healthier and more enjoyable culinary journey.
What it does
Currently, the Nutri AI app takes user's prompt and return a dish recommended to the user, with nutrition facts, budget, and recipe information. User can also download responses and log in to the application. In the future, the Nutri AI app aim to create custom recipes to match your groceries, dietary needs, time, and budget, elevating your daily meals. It also suggests nearby supermarkets for efficient shopping, simplifying meal planning and promoting healthy choices.
How we built it
- Utilized a stack of programming languages and frameworks, including Python and Streamlit for web development.
- Integrated various APIs, including the OpenAI API, DALL-E to access generative AI models for content (text & image) generation.
- Collaborated closely with team members, each contributing their expertise in front-end and back-end development, prompt engineering, and business.
Challenges we ran into
Integration of Front and Back End: Ensuring seamless communication between the user interface and server-side logic, merging technical and user experience challenges.
Prompt Engineering for Recipes: Crafting precise, well-structured recipe prompts through natural language processing and machine learning, considering user preferences.
Time Constraints: The challenge of optimizing workflow and resource allocation to meet project goals within limited time, requiring efficient project management and decision-making.
Accomplishments that we're proud of
Successful Prototype Development: ALthough we have We excelled in prioritizing between features, focusing on the most impactful aspects.We have triumphed in crafting a fully functional prototype website for our Nutral AI Project.
Harnessing Generative AI: Our team harnessed the immense capabilities of Generative AI by employing prompt engineering techniques. This approach enabled us to generate highly valuable information personalized for user's
Feature Integration for User Benefit: We seamlessly integrated a range of user-centric features into our system. Notably, users can not only receive detailed dish information but also access images of the dish, download the recipe, and receive valuable insights on the associated costs in terms of money and time.
User Interface Implementation: We successfully developed a user-friendly interface that allows people to interact with our AI system. This interface includes user login functionality, ensuring a personalized experience for our users.
Cloud Deployment on Azure: We containerized our application. We are prepared to deploy container to the Azure cloud, we've established a robust and reliable platform, assuring a smooth and accessible user experience.
What we learned
Our key takeaway from this project is the skillful utilization of APIs to construct models that harness the power of Generative AI. This newfound expertise has empowered us to craft cutting-edge solutions capable of generating text and images, delivering insightful answers, making valuable recommendations, and much more. Our proficiency in tapping into the potential of Generative AI through APIs has expanded our horizons, equipping us to build intelligent and adaptive applications with the potential to revolutionize a multitude of industries and applications.
Future Project Scope:
In the next phase of our project, we have ambitious plans to enhance and expand our services to provide a more comprehensive and tailored experience to our users. Here are the key areas we aim to focus on:
1. Leveraging RAG and Official Documents: We intend to integrate the RAG (RRetriever-Augmented Generation) model into our platform. Our goal is to utilize this advanced model to train on documents sourced from reputable official websites, such as the US Agriculture Department. This will enable us to provide more accurate and up-to-date information to our users.
2. Customized Responses with User History: To create a more personalized and engaging experience, we plan to harness the power of a user's chat history. By analyzing their past interactions with the system, we can offer responses that are finely tuned to their preferences and needs.
3. Geo-Specific Information: One of our key objectives is to make the user experience more location-centric. We will incorporate geolocation data to provide users with pertinent local information, including real-time food prices, nearby grocery stores, and restaurant recommendations.
4. End-to-End Experience: Our vision is to transform the user journey into a seamless, end-to-end process. This means not only offering assistance with meal planning and recipes but also assisting users in their grocery shopping endeavors. We aim to bridge the gap between planning and cooking, making it a unified and hassle-free experience.
5. Scaling for Wider Accessibility: As we seek to broaden our user base, we understand the importance of technical preparedness. To accommodate increased traffic and user demand, we will focus on building a more robust and secure application infrastructure and database. This will ensure that our platform remains accessible, reliable, and scalable as we extend our reach to more people.
In summary, our future project endeavors revolve around enhancing the quality and personalization of our services, catering to specific user needs, and expanding our platform's accessibility. We are committed to delivering a more integral and user-centric experience, all while ensuring our technical infrastructure is ready to scale with the growing demand.
Log in or sign up for Devpost to join the conversation.