DiaBuddy Health Companion

• 

  Main Dashboard - Real-time health metrics and quick actions

• 

  Save meals by text or ElevenLabs voice, receive Gemini analysis, and get calorie/carbs estimates instantly

• 

  Trends Analysis - Blood sugar patterns and historical data visualization

• 

  PDF Health Report - Comprehensive analysis export for healthcare providers

Inspiration

I built DiaBuddy after talking to a friend with Type 1 diabetes. She described the stress of tracking blood sugar, meals, and medication while trying to live a normal life. I realized people need a companion that listens, understands their patterns, and gives instant advice

What it does

DiaBuddy is a voice-first diabetes companion. Users log blood sugar readings, meals, and medication. The app gives personalized AI insights and generates doctor-ready PDF reports summarizing trends, nutrition, and adherence

How we built it

I used Google Gemini AI for personalized advice, ElevenLabs for natural voice responses, MongoDB to store flexible health records, and Supabase for secure authentication. The frontend is built with React, TypeScript, Tailwind, and Recharts for interactive visualizations. PDFs are generated with jsPDF

Challenges we ran into

Streaming voice without delay, managing API's, syncing MongoDB health records with Supabase authentication, and making the app fully responsive for phones, tablets, and desktops were the toughest problems

Accomplishments that we're proud of

I created a functional, voice-first diabetes companion with real-time insights, responsive visualizations, and doctor-ready reports. It works on multiple devices and transforms raw health data into actionable guidance

What we learned

Healthcare UX must be simple and fast. Context is essential for AI advice. Type safety and structured data prevent dangerous errors. Voice-first design greatly improves accessibility

What's next for DiaBuddy Health Companion

I plan to integrate predictive analytics for glucose trends, real-time CGM syncing, multi-language support, offline-first functionality, and camera-based meal scanning so users can log meals instantly with pictures

Built With

• aihealthinsights
• authentication
• cloudnative
• docker
• elevenlabs
• flask
• googlegemini
• healthanalytics
• healthrecords
• jspdf
• jwt
• meals
• mongodbatlas
• postgresql
• promptengineering
• python311
• radixui
• react18
• reactquery
• realtimeupdates
• recharts
• restapi
• secureapis
• shadcnui
• supabase
• tailwind-css
• tailwindcss
• texttospeech
• timeseriesdata
• typescript
• vite
• voicefirstux