Amma's Health is a family health record that reads my mother's lab reports, remembers them for months, learns from our corrections, and proves every step. My mother is a patient with dozens of lab reports across many visits. Nobody in my family could answer simple questions: is her blood sugar getting worse since March? What did the doctor ask us to watch? What should we ask at the next visit? Amma's Health answers them, and it is built so a family can trust it. Family view: snap a photo of a report and the agent reads every value, compares it with her history, and updates a body map of her health. You see trends against the normal range, every report side by side, a watch list, and questions for the next doctor visit. Each number links back to the original report page it came from. Under the hood: a harness on MongoDB Atlas Agent Engine runs Reader to Verifier to Memory writer to Watcher to Visit prep. It never guesses. A smudged digit pauses the run durably until a family member confirms it, and nothing uncertain is written to her record. It remembers. All four Agent Engine memory types are scoped to the patient: current values, visits with doctor notes, a shared dictionary of test names, and report-reading rules. It learns, with a gate. Your correction becomes a new reading rule, tested against 22 earlier reports with locked answers the agent can't see. It works across weeks. A watcher tracks the doctor's follow-ups over months and flags ones that never happened. It's governed. It never advises treatment: such requests are blocked and rewritten as questions for the doctor. Family members get scoped access (the home aide sees medications only), doctor notes are encrypted, and every held, blocked, denied and rejected action is logged with who did it. Built with MongoDB Atlas, Atlas Agent Engine, Voyage AI, GridFS, change streams, Gemini and Claude.
More from The Harness Engineering & Model Wrangling #SFTechWeek Hackathon
MongoMation is an autonomous PLC engineer. You describe a machine in plain English, and Claude agents write IEC 61131-3 PLC code, compile and deploy it to a…

LabRat is a self-improving harness for a physical lab agent, running on a real bench: two SO-101 arms, a 3018 CNC carrying a servo pipette I designed during…

Recursive Harness Scientist: an AI scientist that improves its own scientific method, and must replicate every gain on data it has never seen. Why it…

Super Bid is a self-improving AI estimator that teaches itself to price bids the way a specific company would. It practices on the company’s real past bids…

A robot that gets better at handling objects it’s never seen, without retraining. After every round, an AI coach reviews what broke and rewrites the robot’s…
WORLDSTATE gives physical spaces a memory. Any phone becomes an observer: open a link, point the camera, and WORLDSTATE tracks what's inside fridges, shelves…
