Notes from inside the clinic inbox
Shorter pieces on the operational detail: what to measure, when to escalate, why the knowledge base decides everything, and what patients actually send at midnight.
6 ARTICLES · 27 MIN OF READING
Operations
No-shows are a messaging problem before they are a policy problem
Most clinics respond to no-shows with a deposit policy. It works, and it also filters out patients you wanted. The messaging fixes come first.
4 min read
Measurement
The five numbers worth tracking on a clinic front desk
Most clinic messaging dashboards report volume, which nobody can act on. Five numbers actually change decisions, and one popular metric quietly causes harm.
4 min read
Language
Arabic, English, and everything in between
Bilingual patient messaging is not a translation problem. It is a detection, terminology and review problem, and getting it wrong is visible to the patient immediately.
5 min read
Design
When an AI should stop and get a human
Handoff is not the failure case of an AI receptionist. It is the feature that makes the rest of it deployable in a clinic at all.
5 min read
Knowledge
Building a knowledge base an AI can actually answer from
Clinics upload a folder of PDFs and wonder why the answers are vague. The problem is almost never the model. It is that the material was written for humans who could ask a follow-up question.
5 min read
Operations
The real cost of the message nobody answered
Every clinic knows it misses after-hours messages. Very few have measured it, and the arithmetic is usually worse than the guess.
4 min read
Reading is cheaper than a pilot
When you want to see it against your own inbox, bring forty real messages and we will run them live.