GlucoBalance

A tool designed for type 2 diabetes prevention, early detection, and ongoing management.

Comment

Inspiration

An old friend has been battling type 2 diabetes for close to two decades. The ailment came with its challenges such as chronic kidney failure and major financial strain. My friend has been candid enough to inform me that she is at risk of developing heart complications. With further research, I noted that millions of people face similar struggles worldwide, many of which could be reduced through early detection, lifestyle changes, and consistent monitoring.

The recent report by the UK's National Institute for Health and Care Excellence reveals that close to 4.6million in the UK are diagnosed with Diabetes of which Diabetes type 2 accounts for 90%. To add on, the report reveals that there is a possibility of around 1.3 people in the UK have undiagnosed type 2 diabetes.

At the same time, global figures highlight a worrying trend. According to the International Diabetes Federation, around 1 in 9 adults are currently living with diabetes, and many remain undiagnosed. Projections suggest that by 2050, this will rise to 1 in 8 adults, this is about 853 million people. More than 90% of these cases will be type 2 diabetes, largely linked to urbanisation, ageing populations, reduced physical activity, and rising obesity levels.

While the numbers are alarming, the impact can be reduced. Preventive measures, early diagnosis, and proper care can help people manage the condition more effectively and avoid or delay serious complications. This inspired me to build GlucoBalance which is an AI-powered companion designed to support prevention, early detection, and ongoing management, giving individuals the tools to take control of their health.

What it does

GlucoBalance is a mobile-first Progressive Web Application (PWA) designed for diabetes prevention, early detection, and ongoing management. It provides:

  1. A WHO/ADA-based risk assessment tool with empathetic AI explanations.
  2. Personalised nutrition planning adapted to local cuisines, with adherence tracking.
  3. A mental health check-in system with daily mood logs, affirmations, and coping strategies.
  4. Progress dashboards to visualise trends in risk scores, nutrition, and emotional well-being.
  5. A clinician-ready report generator for sharing insights with healthcare professionals.
  6. Notifications and reminders that keep users engaged and motivated in their wellness journey.

How I built it

I developed the application using Amazon’s Kiro IDE, which allowed me to quickly set up both the frontend and backend. On the backend, I designed structured tables for Users, Risk Assessments, Nutrition Plans, Mental Health Logs, and Check-ins. This ensured that data was properly stored, linked, and retrievable in real time.

For the frontend, I built GlucoBalance as a Progressive Web App, mobile-first and responsive, so users can access it on any device without needing a separate install. I used React and Tailwind CSS for the UI, keeping it clean, lightweight, and fast-loading.

The dashboard architecture is card-based, which means all features which are; risk status, mood tracker, nutrition snapshot, progress charts is modular. This approach makes it easier to maintain and extend in the future.

For the AI layer, I integrated Gemini AI. Gemini provides personalised insights in three key areas:

  1. Empathetic explanations after risk assessments,
  2. Daily affirmations and coping strategies for mood check-ins,
  3. Lifestyle tips tied to nutrition and progress data.

Challenges I ran into

  1. Structuring the risk assessment logic and ensuring results were medically aligned with WHO/ADA standards.
  2. Designing a user-friendly dashboard that balances data depth with simplicity.
  3. Integrating Gemini AI in a way that felt empathetic and supportive rather than overly clinical.
  4. Time constraints. I learnt about the hackathon on 21st August 2025.

Accomplishments that I am proud of

  1. Building a functional PWA within the limited timeframe.
  2. Successfully integrating AI-generated insights that personalise the user experience.
  3. Creating a multi-feature wellness tool that unites physical health, mental health, and lifestyle management.
  4. Ensuring the app remains accessible with a clean, mobile-first design.

What we learned

  1. The importance of holistic health management, combining prevention, nutrition, and emotional support.
  2. How to leverage Kiro IDE’s capabilities for rapid development and database handling.
  3. Effective integration of AI models into health-focused applications, balancing empathy and accuracy.
  4. The value of iterative design, where user experience plays as critical a role as backend functionality.

What's next for GlucoBalance

I will introduce moderated and safe community features for peer support and group challenges. I will also introduce smartwatch integration for real-time health metrics and multilingual support for global accessibility.

Built With

Share this project:

Updates