AI remote monitoring platform
Your clients' readings, in one record. Longova messages your team and your clients when something needs a look.
Layer 01
Readings from home
Clients agree, on a screen in the app, before any data moves.
The scale and the Pulse band sync on their own. Clients log meals by photo, voice or text, and upload labs as a photo or a file. Care teams add the audio of a visit, and it becomes text in the same record.
Each reading joins a queue the moment it arrives and is checked before it moves on. A reading that fails a check is set aside for review, not lost, and a sync that runs twice never counts a reading twice.
Consent at the gate
Each client agrees to what is shared before the first reading moves. The answer is stored with their record.
Locked on the way in
Every sync and upload travels over a locked connection. Nobody on the network between the phone and Longova can read it.
Layer 02
Locked and kept apart
The models that read a record never learn whose record it is.
Names, contact details and health numbers live in an identity vault. The health data carries a client ID in their place. Your care team’s dashboard is the one screen that shows a name beside the data.
Everything is stored locked, and the keys sit apart from the data they protect. The vault has keys of its own, so opening the health data never opens the names.
A vault for names
Names and contact details sit in a separate store. Only a client ID links the vault to the health data.
Age in, names out
Model calls carry a client ID and an age, never a birth date. Names in notes and transcripts are scrubbed out first.
Layer 03
Record and agent
Each client’s record is a context graph of linked entries.
Every entry comes out in the same shape, with its time, its source, its unit and a standard health code. A meal links to the weigh-in after it, and a poor night links to the heart rate that follows.
The Longova agent reads the whole record, not one reading at a time. It works out trends with fixed math first, then checks them against your team’s rules and reviewed studies before it writes a word.
One shape per entry
Pounds or kilos, and each lab’s own unit, all line up. The original value is kept, so anyone can trace a number to its source.
One record, read by both
The agent and your care team read the same facts, so a missed dose, a side effect and a bad night show up as one story.
Layer 04
Signals
The agent speaks up when a record changes, and no one has to ask.
It wakes on a new reading, a logged dose, a lab upload or a missed weigh-in. Most events simply join the record. When one matters, the agent drafts a signal with the entries that set it off.
Your care team approves drafts before they reach a client. Any claim a study does not support is held back, and the agent never diagnoses, prescribes or changes a plan.
For your care team
Alerts that list their evidence, a brief before each visit, and a short summary of what the client is worried about.
For your client
Plan reminders, nudges that name the next step, and plain answers about their own trends, by push, SMS or chat in the app.