Articles
Notes on how a clinic actually runs, what practice software usually gets wrong about it, and the decisions behind LogBlues. Written by the people building it, for the people using it.
Start here
Workspace
Most clinics run on a scheduler, a chart, a billing tool and four spreadsheets. What changes when they are the same thing.
Read →
Blues
Clinical AI usually works by sending the chart to a third party. We tried building it the other way, and here is the trade.
Read →
Security
One question shapes more of your risk than most, and it rarely comes up in a demo: where do the charts actually sit?
Read →The archive
Nineteen pieces. Pick a subject and the list narrows.
Most clinics run on a scheduler, a chart, a billing tool and four spreadsheets. What changes when they are the same thing.
Every appointment, note, file, case and invoice hangs off one patient record. Getting to it takes one search.
Day, week and provider views, reminders that go out on their own, and a check-in that starts the visit.
Vitals, notes, procedures and codes in the place the work already happens, instead of a separate program afterwards.
Threads attached to the patient they are about, files that travel with them, and nothing routed through a consumer chat app.
Where the scan, the outside record and the signed consent live, who can open them, and how you get one back in a hurry.
Every clinic runs on follow-ups living in somebody’s head. What happens when they get an owner and a due date instead.
Cases group a course of treatment. Referrals come back to a person instead of going quiet for three weeks.
CPT, HCPCS and ICD-10-CM where the work happens, invoices that follow, and balances you can see without running a report.
Ten roles across sixteen domains, deciding which screens and dashboard cards each person reaches. How the permission model is drawn, and why it is drawn there.
Who is coming, what is outstanding, what needs a person today, on one screen before the first patient arrives.
Clinical software is famous for being unpleasant to use. That is a choice somebody made, and it can be made differently.
Clinical AI usually works by sending the chart to a third party. We tried building it the other way, and here is the trade.
One question shapes more of your risk than most, and it rarely comes up in a demo: where do the charts actually sit?
Your waiting room is full and the system is unreachable. What a local-first design changes about that afternoon, and what it does not.
Charting was supposed to save time, and for a lot of clinicians it has not worked out that way. Where the friction tends to sit.
A hospital has analysts. A smaller practice usually has the owner, after hours. The data already exists; something has to surface it.
Larger practices have procurement departments to buy for them. Smaller ones are often left choosing between enterprise pricing and nothing much.
Software usually asks the practice to adapt to it. We tried it the other way round, and named the idea after the inversion.
Nothing on that yet. Write to us and we will fix it.
Try it
Two months free, no card and no sales call. Install it, put a week of your real schedule through it, and decide from there.
Start free Every feature, written down →