LUCID

Overview

LUCID is a phone-based AI assistance prototype designed to help recognize potential head-injury emergencies in low-resource settings. Its design centers on ordinary phone calls, including landlines, so callers would not need to download an app or have mobile internet. The system still relies on backend infrastructure.

Role: Software system builder

Collaboration: Team: Walter Shewmake, David Ross, and Rodney Courtright. I built the software system; David developed the clinical decision logic.

Conversation and decision logic

The prototype design separates conversational AI from an explicit clinical rule set. AI interprets the caller's responses, while the rule set handles decision and escalation logic. My software work connects this flow with persistent conversational context, follow-up calls, and routing; David developed the clinical decision logic.

Prototype design

• 1Multilingual voice interaction and structured questions about potential warning signs
• 2Conversational context retained across interrupted calls
• 3Automated follow-up calls
• 4Escalation and call-routing features when available

Outcome & Status

Our three-person team won first place at the 2026 mission:BRAIN Global Neurosurgery Hackathon. LUCID is a hackathon prototype, not a clinically validated or deployed medical service.