1 Referral Management AI Agent, Live in Production.

Referral intake is a bottleneck built out of incomplete information. A letter arrives, it has to be read against the receiving service’s acceptance criteria, and often it is missing something — a result, a measurement, a confirmation that a prior step was tried. The referral is then rejected or returned, the referrer resubmits weeks later, and the patient waits through both cycles for a reason that was knowable on day one. The check itself is not clinically difficult; it is a comparison between what the letter contains and what the criteria require. But it needs someone to read the whole letter carefully, and that person is usually the same clinician whose triage list is the reason the queue exists.

Referral Intake Review checks an incoming referral letter against the receiving service’s acceptance criteria, extracts the clinical detail, and says plainly whether it can be booked or what is missing before it can. The second half is the output that shortens the pathway: a specific list of what to obtain turns a rejection into a request, which the referrer can act on immediately. One agent covers this process. It does not decide clinical priority or who should be seen — it reports whether the stated criteria are met and what is absent, and the triage decision remains a clinical one.