Resilience
When the software goes down, the clinic shouldn’t
Cloud-hosted practice software turns someone else's outage into your closed clinic; local-first software keeps the record where the care happens.
The waiting room is full and the software is gone
Your clinic is fine. The lights are on, the staff showed up, the patients are in the chairs. But the software that holds every chart, every appointment, and every note lives somewhere else, and right now it is not answering.
It might be your internet connection. It might be a hosting incident hundreds of miles away. It might be a deploy that went out on someone else's Tuesday morning. From inside the clinic, all three look identical: nothing loads, and there is nothing you can do about it.
That is the real cost of putting the record somewhere you cannot reach without a working connection. You are not just renting software. You are borrowing the ability to see your own patients.
Local-first means the record is where you are
LogBlues runs on the computers already in the practice. The record is written there, read there, and backed up there. There is no round trip to a distant server standing between a clinician and a chart.
So a dropped connection is an annoyance, not a closed clinic. The full workspace stays usable offline: charting, scheduling, everything. Email may stall and the weather widget may go stale, but the practice keeps running.
This also changes what a record costs. When storage is local, the price of keeping more history is a disk, not a subscription tier. Retention becomes a hardware decision instead of a billing conversation.
Sync is a feature you turn on, not a condition of use
Cloud sync is not included with the current build. Where it exists it is something a practice switches on when it wants multi-site access or off-site redundancy, and never the thing that makes the software work at all.
When it arrives, it will look like this:
- An authenticated API in front of storage, so the app never talks to a database directly.
- Storage isolated per practice, with tenant isolation enforced at the row level in the database rather than trusted to application code.
- Designed for HIPAA compliance from day one, not retrofitted once a customer asks.
The difference matters on the bad day. If sync is optional, an outage in the sync layer degrades one feature. If sync is the foundation, an outage in the sync layer degrades your entire practice.
Backups you have actually tested
Local storage is only an advantage if the local copy survives. A dead drive in the back office is every bit as bad as an unreachable server, and it is a failure mode that cloud vendors handle quietly on your behalf.
So scheduled backup and tested recovery are part of the Security Center. Not a checkbox that claims backups are enabled, but a routine that runs on a schedule and a restore path you can exercise before you need it. An untested backup is a hope, not a plan.
The assistant stays too
Blues, the assistant in LogBlues, runs on the clinic's own machine. That is an unusual choice right now, and a deliberate one.
Assistants that call out to a hosted model inherit every outage, rate limit, and deprecation upstream. If the provider has a bad day, the feature you built your workflow around simply is not there. Running locally means no upstream AI outage can take Blues away from you.
It also means the clinical context Blues works with never leaves the building to get useful. The privacy benefit and the availability benefit come from the same design decision.
What resilience actually looks like
Nobody chooses practice software for the day everything works. Every option looks fine then. The differences only appear when something breaks, and by then the architecture is already decided.
A local-first system fails in smaller pieces. Lose the internet and you lose email and sync. Lose a machine and you restore from a backup you have already tested. Neither one empties your waiting room.