Skip to content

Beacon

Our goal is to optimize every aspect of the clinical trial workflow for clinician, patient, and pharma stakeholders by extracting insights from voice inputs to increase operational efficiency in enrollment, adverse event

Code Video Cerebral Valley

Core trick

Screens patients against protocol criteria in a reusable workflow, then extracts disqualifying events from voice follow-ups for clinician confirmation.

Tags

healthvoice

Description

Our goal is to optimize every aspect of the clinical trial workflow for clinician, patient, and pharma stakeholders by extracting insights from voice inputs to increase operational efficiency in enrollment, adverse event detection, and clinical workflow support.
developingmedicine/Hackathonia★ 00 forksTypeScriptCreated Jul 18, 2026 · last push Jul 18, 2026

README excerpt

# Beacon — Clinical Trial Enrollment Copilot

**Anthropic Hackathon build · Jul 18, 2026**

> A clinician-guided enrollment copilot that turns protocol criteria, clinician
> expertise, and patient evidence into a **reusable screening workflow** — then
> keeps watching enrolled patients, surfacing disqualifying events from routine
> follow-up conversations before they slip through.

Beacon is decision support, not automation: it screens, explains, and flags —
but a clinician confirms every status change. Nothing is diagnosed or
disqualified autonomously.

---

## The problem

Screening patients for a clinical trial is slow, manual, and lossy:

- **Eligibility is buried** in dense protocols and scattered across the chart —
  BMI trends, ICD-10 codes, labs, vitals, free-text notes.
- **Clinician expertise doesn't persist.** A trial physician knows "watch alcohol
  and biliary disease for pancreatitis risk on a GLP-1" — but that judgment lives
  in their head, re-applied ad hoc, patient by patient.
- **Disqualifying events after enrollment get missed.** A participant mentions a
  hospitalization in a routine check-in; unless someone connects it back to an
  exclusion criterion, they stay enrolled when they shouldn't.

## What Beacon does

Beacon turns a one-time screening effort into a durable, auditable workflow:

1. **Parse the protocol** into structured, evaluable criteria — labeling each as a
   deterministic **Rule** (BMI ≥30, pulse 60–100, ICD-10 exclusions) or a
   **Judge

Team