Diabotes

Inspiration

With over 400 million people worldwide affected by diabetes, there is a pressing need for innovative solutions that can help individuals manage their condition more effectively and conveniently.

We recognize that managing diabetes can be a complex and overwhelming task, requiring constant monitoring, lifestyle adjustments, and informed decision-making. Many individuals struggle to navigate this journey alone, often lacking the support, knowledge, or motivation needed to make lasting changes.

This is where the vision for Diabotes was born – to create a personalized, accessible, and data-driven tool that could bridge the gap between medical advice and daily life. By leveraging the power of wearable technology and artificial intelligence, we saw an opportunity to provide individuals with diabetes with a virtual companion that could offer real-time guidance, support, and motivation tailored to their unique needs.

Our team's goal is to empower individuals with diabetes to take control of their health and improve their quality of life. We believe that by harnessing the latest advancements in AI and wearable technology, we can revolutionize diabetes management and make a meaningful difference in the lives of millions.

What it does

Introducing Diabotes, the revolutionary chatbot that empowers individuals with prediabetes and diabetes to take control of their health like never before. By seamlessly integrating with your wearable devices, Diabotes harnesses the power of real-time data to provide personalized, actionable insights for managing your diabetes effectively.

Diabotes is an AI-powered chatbot designed to transform the daily management of diabetes into a more intuitive, personalized, and less overwhelming experience. Leveraging real-time health data inputs—like continuous glucose monitoring, heart rate, and sleep patterns—Diabotes provides users with instant, expert advice on nutrition, exercise, and lifestyle adjustments tailored specifically to their unique health profiles and daily routines. Unlike traditional health management tools, Diabotes offers a conversational, supportive experience that feels like chatting with a knowledgeable friend who's always there, day or night, to help navigate the complexities of diabetes care.

Our platform stands out by prioritizing the user experience, making diabetes management more accessible and less clinical. It's designed to validate and support the user's efforts, offer practical solutions, and encourage a proactive approach to health.

Whether you need guidance on carb counting, meal planning, or finding the motivation to stay active, Diabotes is here to support you every step of the way towards a brighter, healthier future.

How we built it

We used Modal to deploy a server bot written in Python and accessible through Poe. The server bot makes real time API function calls to Vital (tryvital.io), which provides API keys for access to a wide variety of wearable devices. We integrated devices like a Garmin watch for heart rate, Oura ring for sleep score, and Abbott Continuous Glucose Management (CGM) device for blood sugar response. Thus, the Diabotes Poe bot has real time access to wearable data! In its system prompt, Diabotes has a RAG to knowledge base including major publications in dietary science and medicine including: The Art and Science of Low Carbohydrate Living, American Diabetes Association: What Can I Eat? (Smart Snacks), and Diabetes Technology: Standards of Medical Care in Diabetes—2022.

Challenges we ran into

Time constraint only! Everything else was smoothly developed and deployed!

Accomplishments that we're proud of

Deploying a Poe server bot with Modal; function calling Vital API; Vital integration with Garmin, Oura, and Abbott CGM; Using Vital sdk; RAG to dietary/medical knowledge base.

What we learned

Poe bots can access real-time data from wearables and other integrations. The bot we developed can provide extremely valuable and informed dietary recommendations for managing health .

What's next for Diabotes

Our current implementation uses the team members' own wearable devices. In the future, we aim to make the chatbot personalized so the users can connect their own devices. We also want to implement Brave Search AI and user GPS location data to find diabetes-friendly restaurants nearby. Restaurant menu parsing for determining diabetes-friendly choices is also in the works.