Health Companion

Health Companion is a voice-powered app for tracking your mental and physical well-being. Log daily updates, detect mood, track symptoms, and get personalized insights—all hands-free and stress-free.

Comment

🩺 Health Companion: A Voice-Powered Health Journal

🌟 Inspiration: Turning Words into Wellness

The journey of building the Health Companion app began with a simple yet powerful realization: keeping track of how we feel is harder than it should be. Whether managing chronic conditions, mental well-being, or general health, patients often struggle to record their symptoms, emotions, and experiences.

For many, the task of journaling feels like an extra burden, especially when health itself is already a challenge. This struggle becomes even more pronounced for those who are not native English speakers, facing both language and cognitive barriers when trying to explain their condition to healthcare providers.

💡 What if simply talking could be enough?
That’s how the idea for Health Companion was born—a voice-powered app that allows users to log their well-being by simply speaking. No typing, no complicated forms—just a conversation with an empathetic AI, always ready to listen.

But this project aimed to go further. It wasn't just about personal tracking; it was about empowering better communication with doctors. Imagine walking into an appointment with a detailed health summary, not from memory but from consistent daily logs. That’s the future Health Companion seeks to enable.

⚙️ How It Works: A Seamless Ecosystem

Building Health Companion required integrating several powerful platforms into one cohesive solution:

  1. ElevenLabs Conversational AI:

    • The heart of the project. This intelligent agent listens, transcribes, and analyzes user speech, extracting moods, symptoms, and general health insights.

  2. Typescript App (Vite + ElevenLabs Module):

    • A user-friendly web interface where users can interact with the agent, view their logs, and track their health trends.

  3. Supabase (Authentication + Database):

    • Secure storage for user data, ensuring privacy while enabling real-time access to health logs.

The process is simple:

  1. 🎙️ Speak: Users share how they're feeling.
  2. 🤖 Process: The ElevenLabs agent transcribes and extracts key health data.
  3. 📊 Store: The entry is securely saved in the Supabase database.
  4. 📅 Review: Users can revisit their logs, and doctors can access summaries for informed consultations.

🚧 Challenges Faced: From Idea to Reality

Like any ambitious project, bringing Health Companion to life wasn’t without its hurdles. Each challenge turned into a lesson, deepening understanding and refining the solution.

🔄 1. Bridging the Agent and Supabase Database

The first significant challenge was ensuring the agent’s logs made their way into the Supabase database. While the agent could transcribe speech and extract insights, getting that data into the database involved complex communication between APIs.

  • Problem: Logs weren’t appearing in Supabase despite successful conversations.
  • Solution: Careful debugging revealed the need for properly authenticated API requests and Row-Level Security (RLS) policies in Supabase to ensure each user could only access their own data.

🕵️ 2. Tracking Errors in Communication

Even after establishing the API pipeline, intermittent errors persisted. Logs would sometimes fail without clear explanations. Diving into Supabase logs revealed 403 (Forbidden) and 400 (Bad Request) errors.

  • Problem: Misaligned data structures and inconsistent API headers.
  • Solution: Implementing detailed error handling and logging in the agent allowed for real-time insights into what was going wrong. It turned out missing user IDs were causing silent failures.

📦 3. Troubleshooting the JavaScript Module

When integrating the ElevenLabs Vite module in the Typescript app, unexpected issues arose. Some voice inputs weren’t processed, and the app occasionally froze.

  • Problem: Missing environment variables and incomplete module configuration.
  • Solution: Thoroughly reviewing the ElevenLabs documentation uncovered the issue: API keys and project IDs were not fully passed during initialization. Proper .env setup resolved the problem.

🎓 Lessons Learned: Beyond Code

While the technical challenges were significant, the true takeaway was the depth of learning about platforms, integrations, and system design.

  1. ElevenLabs:

    • Understanding conversational AI beyond simple transcription—voice tone, context awareness, and multi-language support.

  2. Supabase:

    • Realizing the power of PostgreSQL in a serverless environment and the importance of Row-Level Security (RLS) for user-specific data protection.

  3. Typescript & Vite:

    • Learning how environment variables and build-time configurations can make or break an app.

  4. API Debugging:

    • Mastering the art of API error tracking, from network logs to Supabase audit trails.

🚀 The Final Impact: Health Tracking Made Simple

Today, Health Companion is more than just a health tracker—it's a bridge between patients and their healthcare providers. It allows users to:

  • 🎙️ Log health updates effortlessly through conversation.
  • 🌎 Communicate across language barriers with multilingual support.
  • 📈 Share structured reports with doctors for better consultations.
  • 🔒 Keep their data private, secure, and accessible anytime.

For users, it’s peace of mind. For doctors, it’s clarity. And for developers, it’s proof that even the most complex ideas can come to life with persistence and the right tools.

👥 Team Member

This project was built by Juan C Rojas, driven by the passion to make health tracking effortless, accessible, and meaningful.

💡 Looking Ahead: The journey doesn't end here. Future iterations will focus on predictive insights, personalized recommendations, and even deeper integration with healthcare platforms.

👉 Health Companion isn't just an app—it's a testament to how technology can transform care, one conversation at a time.

Built With

  • conversationalai
  • elevenlabs
  • lovable
  • supabase
  • typescript
Share this project:

Updates