-
Healix AI: An AI-powered health intelligence system delivering intelligent diagnosis with a human-first approach.
-
Users enter symptoms through a simple interface to receive AI-driven health insights and predictions.
-
Gemini 3–powered conversational assistant providing explainable, context-aware health guidance.
-
System architecture integrating ML models with Gemini 3 for explainable and responsible health intelligence.
-
Symptom-based disease prediction with confidence scoring and a structured 360° recovery roadmap.
Inspiration
The inspiration for Healix AI came from the realization that while the internet is flooded with medical information, it is often scattered or overwhelming. During the post-pandemic era, the gap between "Googling symptoms" and consulting a doctor became glaringly obvious. We wanted to build a bridge: a reliable, AI-first tool that doesn't just name a disease but acts as a compassionate first responder—offering immediate, structured, and actionable recovery advice.
What it does
Healix AI is a comprehensive health companion that serves three main functions:
Multimodal Diagnosis: Using Gemini 3 Pro, users can upload photos of symptoms or medical reports for instant, high-accuracy reasoning.
Holistic Recovery Roadmap: It generates a personalized 4-part recovery plan:
Precautions: Immediate steps to take or avoid.
Medications: Common pharmacological treatments.
Diet: Nutritional advice to boost immunity.
Workouts: Safe physical activities suitable for the condition.
Conversational Support: A medical chatbot allows users to ask follow-up questions in plain English, providing context-aware answers.
How we built it
We embraced the "Vibe Coding" philosophy by building the core intelligence within Google AI Studio:
Generative AI Engine: We integrated the Gemini 3 Pro model. We chose this for its native multimodality and advanced reasoning capabilities.
Backend: The core application is built on Python Flask. It handles routing and retrieves data from structured medical datasets.
Frontend: We used a responsive HTML/CSS interface with a dark-mode design to ensure ease of use during stressful health moments.
Challenges we ran into
Prompt Precision: Ensuring the AI remains grounded in medical facts. We solved this by implementing strict system instructions in AI Studio to minimize hallucinations.
Data Mapping: Mapping predicted labels to corresponding rows in auxiliary datasets required precise string matching and data cleaning to ensure no "key errors" occurred during runtime.
Accomplishments that we're proud of
Hybrid Intelligence: Successfully merging deterministic data processing with probabilistic generative AI into a single workflow.
Native Multimodality: Implementing a system that can "see" symptoms and medical documents to provide better context than text alone.
What we learned
The Power of Reasoning: We learned that Gemini 3’s ability to reason through health history is more valuable than simple database searches.
Deployment is a Skill: Handling environment variables and cloud constraints taught us that "it works on my machine" is only the beginning.
What's next for Healix AI
Wearable Sync: Integrating with APIs to use real-time vitals like heart rate.
Doctor Connect: Adding a feature to locate and book appointments with nearby specialists.
Log in or sign up for Devpost to join the conversation.