Parent. Technologist. Rare disease advocate. Two decades in geospatial intelligence. One decade building infrastructure for a daughter who communicates in ways the world hasn't learned to receive yet.
My daughter was diagnosed with DDX3X syndrome — one of the leading genetic contributors to intellectual disability in females globally, yet still significantly underdiagnosed. She is nonverbal, and like every nonverbal individual with a complex, multi-system profile, she spent years in rooms where the people responsible for her care each understood one piece of her. Nobody held all of it. Nobody had the time, the data, or the infrastructure to hold eleven years of her simultaneously in mind.
I watched decisions get made about her life from incomplete pictures. I watched a misdiagnosis enter her record and stay there for years, shaping how clinicians understood her visual profile. I watched services go unclaimed because nobody had the bandwidth to find them. I watched caregiver transitions — each one a small loss of accumulated knowledge that existed nowhere except in someone's memory.
The tools to solve this didn't exist. So I built them.
"The gap is not in the person. It is in the infrastructure around them. I decided to build the infrastructure."
HerVoice is a private, AI-assisted longitudinal care intelligence system — built on a dedicated home server, never cloud-connected, capturing structured observations from every member of a care team and synthesizing them into a continuously updated portrait of who an individual is right now. It is the living version of a Letter of Intent. It is the system that doesn't forget, doesn't retire, and doesn't move on to the next patient.
The architecture is built for three phases: today, where it solves the coordination and continuity crisis facing nonverbal individuals and their families; near term, where AI trained on years of longitudinal individual data closes the gap between what a person understands and what they can express; and the future, where the richly characterized individual model becomes the foundation that makes robotic caregiving and ambient AI assistance genuinely useful — because the technology will finally have someone to work from.
Whether you are a family navigating this same challenge, a researcher or clinician working in rare disease, a technologist who sees what we see, or an organization looking to connect — the conversation matters.
Connect With Us View the Project