Skip to content

Winner at RAISE YOUR HACK Paris Hackathon, VHeal AI is an Agentic AI solution that streamlines hospital discharge by automating summaries, follow-ups, and patient communication with empathy and precision.

License

Notifications You must be signed in to change notification settings

codewithEshaYoutube/Vheal_AI

Repository files navigation

🏥 VHeal AI – Agentic Discharge Copilot for Hospitals

🧠 Raise Your Hack 2025 – Paris (Online + Onsite)
🚀 Built with Agentic AI to automate hospital discharge coordination in real time.
❌ No more delays. ✅ More beds, better care, less burnout.

🔗 Live Demo: https://vheal-ai.streamlit.app


🚨 The Problem: Discharge is Broken

In major hospitals (e.g. Manhattan, Paris), discharge delays are common due to:

  • Fragmented handoffs between doctors, nurses, pharmacy, billing
  • Manual summaries and outdated checklists
  • No centralized platform to track readiness or progress

These issues result in:

  • ❌ Bed Shortages
  • 😫 Staff Burnout
  • 💸 Higher Patient Costs
  • ⏳ Inefficient Care Delivery

💡 Our Solution: VHeal AI

VHeal AI is an Agentic AI-powered autonomous discharge assistant that acts like a digital hospital staff member.

👇 What It Does:

  • 📍 Detects when patients are ready for discharge
  • 📋 Plans and delegates discharge tasks automatically
  • 🔗 Communicates across departments via APIs & messaging
  • 📊 Displays progress on a real-time dashboard

"Not reactive. Not rule-based. Truly autonomous."


🧠 What Makes It Agentic?

🔍 Capability ✅ Description
Trigger-Based Autonomy Starts when a patient is marked "Ready for Discharge"
Multi-Step Reasoning Validates discharge: vitals, meds, labs, signatures
Delegated Communication Notifies departments like billing, transport, pharmacy
Tool Use & Memory Uses EMR, staff calendars, previous discharge summaries
Multimodal Interaction Operates via dashboard, APIs, WhatsApp, and email

🔧 Core Features

  • 🧾 Auto-generated summaries (GPT-4o, Claude 3, Mistral)
  • Checklist Agent – Labs, meds, vitals
  • 📞 Smart Alerts – WhatsApp, Email, SMS to departments
  • 📆 Discharge Scheduling – Room + transport availability
  • 📊 Live Dashboard – Tracks every agent and staff action
  • 🧠 Planning Chains – CrewAI + LangChain multi-agent flow

🖼️ Product Snapshots

📊 Agent Dashboard

Dashboard


📈 Discharge Analytics View

Analytics


💡 Features Overview

Features


🛠️ Tech Stack

Layer Tools
LLMs & Agents GPT-4o, Claude 3, Mistral, LangChain, CrewAI
Backend FastAPI / Node.js
Frontend React.js / Streamlit
Database MongoDB / Firebase
Messaging Twilio, WhatsApp Business API
RAG Layer Vectara / Weaviate (optional)

👥 Team VHeal – [LabLab.ai IDs]

Name ID Role
Eesha Tariq eesha_tariq899 AI Engineer – Agent Logic, LLMs
Aroob Mushtaq aroobmushtaq818 Backend Dev – API Integration
Arfa Tariq Arfah_t Researcher – Clinical Workflows
Zeeshan Tariq zeeshantariqpkn Data Scientist – RAG & Planning
Waqar Ahmed WaqarAhmed555 Frontend Engineer – Dashboards
Sujal Bedre sujal_bedre114 Prompt Engineer – LLM Pipelines

🎯 Hackathon Goal – Raise Your Hack, Paris

  • ⏳ Automate hospital discharge in <72 hours
  • 🧠 Deploy real-time Agentic AI
  • 🔌 Integrate with APIs like Twilio, Firebase, and EMRs
  • 📊 Present live dashboard + action-based workflows
  • 💡 Show the future of autonomous care coordination

“We’re not just solving for hospitals — we’re solving for time, care, and cost.”


▶️ How to Run Locally

# 1. Clone the repository
git clone https://github.com/codewithEshaYoutube/Vheal_AI.git
cd Vheal_AI

# 2. Backend setup
cd backend
pip install -r requirements.txt  # or use npm install for Node.js

# 3. Frontend setup
cd ../frontend
npm install
npm run dev

# 4. Add environment variables (.env)
OPENAI_API_KEY=...
TWILIO_API_KEY=...
FIREBASE_CONFIG=...

About

Winner at RAISE YOUR HACK Paris Hackathon, VHeal AI is an Agentic AI solution that streamlines hospital discharge by automating summaries, follow-ups, and patient communication with empathy and precision.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages