🏥 ICUthere

⚠️ Over 130 million people visit U.S. emergency rooms each year, yet many face critical delays due to manual triage and staff overload. ICUthere uses AI to streamline emergency care and prioritize patients faster.

🧠 Background

ER teams juggle life-or-death decisions under intense pressure, often relying on outdated or manual triage processes. This leads to longer wait times, inconsistent prioritization, and staff burnout.

ICUthere is an AI-assisted triage platform that analyzes patient inputs — like speech, facial expression, and symptoms — to recommend urgency levels and surface high-risk cases instantly. It empowers healthcare staff with faster decisions, improved care, and reduced legal exposure.

💡 What is ICUthere?

ICUthere enhances ER triage through:

  • 🧾 Video + Audio Patient Intake – Patients record symptoms via a friendly voice interface
  • 📊 Triage Dashboard – View patients sorted by severity with AI-generated recommendations
  • 🚨 Urgency Flags – Instantly highlights critical red-flag symptoms
  • 👩‍⚕️ Staff Coordination Panel – Tracks who's on shift and current case assignments
  • 🔁 Real-Time Queue – Dynamically updates as new cases come in

✨ Essential Features

  • 🤖 AI Triage Engine – Combines voice tone, symptoms, and expressions to recommend care levels
  • 🗣️ Empathetic Voice Assistant – Guides patients through check-in calmly and naturally
  • 🔄 Live Patient Queue – Sorted by urgency and synced in real time
  • 🧘 Burnout Reduction Tools – Automates repetitive intake questions for staff relief
  • 🔌 EHR-Ready Integration – Built for compatibility with hospital systems

📚 Takeaways

🧠 What We Learned

Designing AI for emergency care requires a deep balance between technical precision and emotional sensitivity. We explored how empathy-driven voice tech can ease patient anxiety while supporting medical staff under pressure.

🏆 Accomplishments

  • Real-time triage powered by AI inputs
  • Emotion-aware voice assistant built with Hume
  • Seamless, intuitive dashboard for ER decision-making

🔮 What's Next

  • Expand symptom training data
  • Add multilingual support
  • Enhance AI transparency for medical staff
  • Improve HIPAA compliance and patient data security

Built With

  • nextjs
  • shadcn
  • tailwind
Share this project:

Updates