A caregiver's second brain
A personal health record that does its own data entry, and keeps a knowledge base that stays current as the research changes.
We set up a private server for your family. Records, questions and research stay on it.
The job nobody assigned
Somebody in every family keeps the shoebox.
The discharge summaries. The lab printouts. The pill bottles photographed in a hospital corridor because you will not remember the dose by the time you get to the car.
AARP and the National Alliance for Caregiving, Caregiving in the US 2025.
Shared medical records
It didn't store the file. It read it.
Upload what your doctor gave you. The agent reads it, structures it, and files every value into the right table for the right family member: nine panels out of one seven-page PDF, each result carrying its own value, unit, reference range and abnormal flag.
Sixty-eight results went in. Two came back as plain sentences with the number and the reference beside them, and every value becomes a line you can chart.
| Test | cholesterol |
| Panel | lipid panel |
| Value | 225 mg/dL |
| Reference | < 200 |
| Flag | high |
| Observed | 27 Jul 2026 |
And the part that happens every day
Say it the way you'd say it to a nurse. Three untidy sentences become a weight, a blood pressure, a heart rate and an exercise entry, each with a timestamp.
Research together
Name a condition. Create a knowledge base.
A knowledge base the agent writes, revises and consults, built to hold evidence that conflicts and changes.
A traditional knowledge base wants one settled answer per question. Medical research does not supply one. Findings conflict, get qualified, and get superseded. So the page is not a static article. You point at what is missing or wrong, and the agent rewrites that section, holding the qualification rather than flattening it.
The pages are not filed away to be read later. They are the context the agent reads before it answers, and you choose which ones count. Ask "do I have risk for heart disease?" and the answer is built from this family member's own results plus the pages you ticked. Not from the open internet.
How it's deployed
One family, one computer
Every family gets its own agent, deployed on its own machine: a VPS, or a physical computer in the house. The medical records and the knowledge base stay there.
Deployed per family, not a shared account on a shared system.
A VPS we set up for you, or a physical computer you keep at home.
The record and the knowledge base live on that machine, private to your family.
Questions caregivers ask
Where does our data actually live?
On your family's own machine: a VPS we set up for you, or a physical computer you keep at home. The records and the knowledge base stay there.
Who can see it?
The people you invite. It is your own agent, deployed per family, not a shared account on a shared system.
What can I upload?
Bring the messiest PDF your hospital ever sent you. That is the one we want to read. The agent reads it, structures it, and files every value against the right family member.
Do I have to type anything in?
No. OCR plus an agent that understands what it is reading does the job the outsourced keyboard did, in the time it takes to make tea.
Does the research come from the open internet?
No. Answers are built from that family member's own results plus the knowledge pages you have ticked as context.
What happens in the first month?
We deploy your private agent computer and help you load the first records. The first month is free, and you can cancel at any time.