DEMO VIDEO
PROBLEM STATEMENT:
Emergency departments face a systemic problem: life-or-death triage decisions occur in chaotic, overcrowded settings where emotion, cognitive overload, and bias can undermine accuracy and fairness. In Canada, patients can wait dozens of hours for beds or assessment, revealing severe bottlenecks and unsafe delays in care. Studies show minority patients are more likely to receive lower-acuity triage scores and experience longer waits, indicating entrenched inequities in current systems. At the same time, clinicians spend substantial time on documentation and electronic records, further reducing capacity for rapid, consistent triage and timely treatment.
SOLUTION:
Our solution is the ER Triage System Gemini Assistant, an agentic AI copilot that automates and optimizes emergency triage to reduce delays, bias, and cognitive load on staff. It ingests multimodal inputs such as vitals, symptoms, notes, and historical data, then generates evidence‑backed triage classifications and risk scores in real time, surfacing the sickest patients first. The assistant integrates with existing hospital systems and EHRs to auto-draft documentation, update triage levels dynamically, and trigger alerts or workflows for critical cases, shortening time‑to‑care while keeping clinicians in the loop for final decisions.
