MedCall AI - Automated Emergency Resource Verification Agent

Inspiration## 💡 Inspiration

During critical medical emergencies, finding vital healthcare resources like ICU beds, ventilators, or specialized units requires making dozens of frantic phone calls to hospital desks. This manual calling bottleneck costs precious, life-saving minutes. We built MedCall AI to eliminate this friction by deploying autonomous AI voice agents that coordinate and verify hospital resource availability in real time.

🚑 What it does

MedCall AI is an emergency resource coordination platform powered by CALL-E Voice Agents:

• Autonomous Inquiries: Triggers automated, natural-sounding phone calls to hospital reception desks to check bed availability for routine emergency planning.
• Real-time Live Tracker: Provides emergency coordinators with an interactive dashboard that reflects live verification status (IDLE ➔ CALLING... ➔ AVAILABLE ✅).
• Structured Intelligence: Parses conversation outcomes and extracts clean status data without manual human intervention.

🛠️ How we built it

• Voice Agent Engine: Integrated the CALL-E SDK (@call-e/calle) with fine-tuned, safety-compliant conversational tasks for voice calls.
• Frontend Dashboard: Built an interactive, dark-mode user interface using Next.js (App Router), TypeScript, and Tailwind CSS.
• Backend Pipeline: Developed custom Node.js automation logic to execute calls via CALL-E APIs and process task completion statuses.

🧗 Challenges we ran into

• Safety & Prompt Policy Alignment: Ensuring the voice agent's conversational task politely inquired about general hospital facility availability without violating sensitive emergency policy filters (422 non-emergency inquiry guidelines).
• Real-time State Management: Connecting asynchronous phone call outcomes seamlessly with the live dashboard UI.

🏆 Accomplishments that we're proud of

• Successfully completing live, bidirectional voice calls through CALL-E and capturing real-time call states.
• Building a clean, end-to-end working MVP capable of reducing hours of medical coordination into automated seconds.

📚 What we learned

• Harnessing low-latency voice AI agents for critical real-world workflows.
• Integrating and configuring CALL-E voice tasks and managing conversational agent responses effectively.

🚀 What's next for MedCall AI

• Multi-hospital concurrent calling to discover nearest available ICU beds in parallel.
• Webhook ingestion for deep transcript extraction and automated patient routing. *

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for MedCall AI - Automated Emergency Resource Verification Agent

Built With

• call-e-voice-sdk
• next.js
• node.js
• tailwind-css
• typescript