🌸 Advora: Your Personal Healthcare Advocate 🌸

About the Project

Inspiration

Every woman knows the feeling: you're sitting in a doctor's office, trying to remember when your symptoms started, how severe they were, and what patterns emerged over the past few months. You're met with skepticism. Your concerns are minimized.

Studies confirm what women have long known: serious medical conditions are 66% more likely to be dismissed as "psychosomatic" in women than in men (Misdiagnosis in Women’s Health: A Growing Epidemic). Women wait longer for diagnoses, don’t receive adequate pain alleviation, and their symptoms get dismissed as overreactions.

This is why we wanted to create a tool that allows you to track and analyze trends in your symptoms. This way, you can gain confidence by advocating for yourself at appointments.

Our team, composed entirely of women, drew from our own experiences and those of women around us. We felt that tracking symptoms shouldn't feel like a chore, and the insights derived from that data shouldn't be hidden in spreadsheets; they should be transformed into actionable intelligence that helps women support their claims.

What We Learned

Building Advora was a journey of discovery across multiple domains:

Technical Growth:

Full-stack integration: We learned to seamlessly connect a Flask backend with a dynamic JavaScript frontend, creating a smooth, app-like experience on in the webbrowser.
AI integration: Working with Google's Gemini API taught us how to structure prompts for meaningful health insights and pattern recognition. We discovered that combining structured data (symptom tags, pain scales) with unstructured input (voice memos, free text) dramatically improves AI analysis quality.
Voice technology: Implementing ElevenLabs speech-to-text opened our eyes to the power of voice interfaces--sometimes it's easier to speak how you're feeling than to type it.
Database design: Structuring our Supabase schema taught us to balance flexibility (for diverse symptoms) with query efficiency (for pattern detection).

Design Insights:

Visual language matters: The severity gradient calendar wasn't just aesthetic--it instantly communicates health patterns at a glance.
Women-centered design: Every decision, from color palette to symptom options, was made with women's experiences in mind.

Key Features Implemented:

Smart Symptom Logging: Users can select from 9 common symptoms, rate pain on a 1-10 scale, add free-text notes, and record voice memos, all in one intuitive interface.
Health Factors Tracking: Period status, birth control type, sickness, and stress levels provide crucial context for AI analysis.
Visual Calendar: Days are color-coded by severity (light pink for mild, deep purple for severe), with symptom icons for at-a-glance understanding.
AI-Powered Insights: Google Gemini analyzes patterns, identifies correlations with period cycles, flags potential red flags, and generates a personalized Doctor Discussion Guide.

Challenges We Faced

Database Integration: We were new to using Supabase so we had to figure out how to integrate it with our frontend.
API integration: Similarly, we had to learn how integrate Gemini API and have it mesh well with our current implementation.
Environment: Working with python, git, dealing with file system issues, etc. These were all intricate technical issues that we had to overcome as they popped up.

What's Next for Advora

Community features: Anonymous aggregated insights (with privacy safeguards)
Wearable integration: Connect with Apple Health, Fitbit, and other devices
Expanded symptom library: More options for diverse health experiences
Telehealth integration: Direct sharing of Doctor Discussion Guides with providers