Skip to content

yessasvini23/MedSync_AI

Repository files navigation

build the future gemini google ai studio ai for good status

🚑 MedSync AI

AI-Powered Emergency Triage & Hospital Routing with Gemini

MedSync AI is an AI-assisted emergency coordination platform that helps paramedics and hospitals triage patients, assess risk, and route them to the most prepared hospital in real time — reducing ER overload, improving response time, and saving lives.

Built using Google AI Studio + Gemini as part of Google Startup School: Prompt to Prototype and submitted to the Build the Future Showcase.


🚨 Why MedSync AI Exists

Emergency departments are overwhelmed.
Critical patients wait too long.
Ambulances are sent to overloaded hospitals.
Clinicians make life-or-death decisions without real-time system-wide insight.

MedSync AI was created to answer one urgent question:

What if emergency teams could see patient risk and hospital readiness at the same time?


💡 What MedSync AI Does

MedSync AI provides AI-assisted decision support by:

  • Assigning emergency triage levels
  • Estimating risk of deterioration
  • Recommending the best hospital based on capacity and specialists
  • Triggering time-critical alerts for cardiac, stroke, and trauma cases
  • Producing clear, explainable outputs for human review

The system is designed to support medical professionals, not replace them.


🧠 How It Works

MedSync AI is powered by a Golden Prompt built in Google AI Studio.

Gemini acts as a virtual emergency coordinator, reasoning across:

  • Patient vitals and symptoms
  • Hospital capacity and specialists
  • Travel time and urgency
  • Safety and escalation rules

All system logic lives inside the prompt itself, making the AI the core decision engine.


🧩 System Architecture

[Paramedic / User]
        │
        ▼
[AI Studio Web App or Swift UI]
        │
        ▼
[Golden Prompt (System Instruction)]
        │
        ▼
[Gemini Reasoning Engine]
        │
        ▼
[Triage + Risk + Routing]
        │
        ▼
[Explainable JSON Output]
        │
        ▼
[Human Review & Action]


✨ What Makes MedSync AI Unique
Prompt-as-Logic Architecture — the system instruction is the application
Protocol-Aware Reasoning — safety rules embedded directly in the prompt
Explainable AI — every decision includes reasoning
Disaster-Ready — adapts to mass-casualty scenarios
No Fine-Tuning — built entirely with structured prompting

MedSync AI shows how prompt engineering can replace traditional backend logic for safety-critical systems.


## 🎥 Live Demo & Prototype

### 🎬 Demo Video (≤ 3 minutes)
https://youtu.be/VUoPJCJl3p0

### 🌐 Live App (Google AI / Cloud Run)
https://medsync-ai.streamlit.app/

### 📱 Mobile UI Preview (Swift Triage)
https://swift-triage-ace.lovable.app


###🛠 Tech Stack
Google AI Studio – prompt design, testing, and live sharing
Gemini – AI reasoning for triage, risk, and routing
Prompt Engineering – system-level logic and safety rules
Synthetic Healthcare Data – safe simulation of patient and hospital data


###👥 Who It’s Built For
Paramedics and EMTs
Emergency department clinicians
Hospital operations teams
Public health and emergency coordinators
MedSync AI reduces cognitive load while improving speed and safety.


###⚠️ Safety & Ethics
This project provides AI-assisted decision support only.
It does not replace professional medical judgment or diagnosis.


###🌍 Vision
With real hospital integrations, MedSync AI could become critical emergency infrastructure, helping cities reduce wait times, optimize hospital capacity, and save lives at scale.

###👩‍💻 Author
Yessasvini Sudarshanam
AI & Social Impact Builder

###📜 License
Shared for educational and demonstration purposes as part of the Build the Future Showcase.


Releases

Packages

Contributors

Languages