Referral Pulse — Referral Leakage & Growth Intelligence

What it does

Referral Pulse turns de-identified EMR exports into referral network intelligence: who is sending volume in, where volume is leaking out, which relationships are drifting, and who to contact next.

Referral networks decay quietly. Most practices only notice after volume drops, because the signal is buried in messy exports and manual spreadsheets nobody has time to reconcile.

The loop

• Export from the EMR, upload the CSV, auto-map the columns, review the result.
• Refresh monthly, which is the cadence that catches drift while it is still fixable.
• No EMR integration required for a pilot — the whole onboarding path is CSV.

What you get

• Referrer health scoring on a four-state scale — stable, watch, at-risk, critical — with churn detection.
• Outbound leakage visibility with capture-opportunity flags, alongside inbound source attribution.
• Revenue-weighted relationship prioritization, labeled explicitly as baseline estimates rather than actuals.
• A network graph with drill-down, plus a provider-level view separating inbound from outbound.
• Results framed as Findings, Impression, Plan — structure borrowed from radiology reporting, so the output ends in an action rather than a chart.
• Outreach draft generation for the referrers worth contacting first.
• Dedup and merge tooling for messy procedure and provider naming, which is the real data-quality blocker in raw exports.

Data handling

Designed for de-identified workflows. Do not upload patient identifiers.

This is practice operations analytics. It has no clinical function and makes no clinical claim.

Written and reviewed by Pouyan Golshani, MD — February 2026