Personal Voice Companion System
Patent Pending Technology
A complete voice system, not a medical box. Built for one person at a time. Designed for the world to hear them.
Not sure where to start?
Find the right fit in sixty seconds.
Watch the film
The companion inside every Grand Voice OS kit. Press play, sound on.
When a person loses their voice, they should never lose their voice.
The laugh a family knows in the dark. The way they say a name. The sound that proves they are still in the room. Speech can be taken. We built this so the voice is kept.
Our Story
Grand Voice OS began with one man. A grandfather, a storyteller, the loudest laugh in the room, slowly losing his voice to ALS while every thought stayed perfectly clear. His family did not need a machine that spoke at them. They needed him. His words, his timing, his sound.
So we built the system we could not find: one that banks a person's real voice while it is still easy, adapts as their needs change, and arrives with the family already taught how to listen. No cloud. No stranger's voice. Offline by design, online by desire.
Every kit is assembled for one specific human being. That is not a production constraint. It is the whole point.
The founder film. Ninety seconds on why this exists. (Final cut in production.)
The problem we refuse
or more. That is the legacy price of a voice.
Months of insurance fights for a device that costs as much as a used car. And when the voice finally comes out, it is a stranger's. Flat. Mechanical. Not theirs.
And here is the part nobody says out loud: a locked box is antiquated the day it ships. No upgrade path, no growth. The person keeps changing; the device never does. Year five looks exactly like day one, except the warranty is gone.
Families call it the second grief: losing the sound of someone who is still alive. Then the manual closes, nobody teaches the family how to talk with it, and too many of these devices end up in a drawer.
One complete kit, one honest price, delivered ready. The voice inside it is banked from the person themselves, in their own sound, while it is still easy to do.
One minute of their voice is enough. A saved voicemail, a birthday video, an old message. From under sixty seconds of real speech, the system rebuilds a voice the family recognizes, and every phrase banked afterward makes it stronger.
And if the voice is already gone, or there was never a recording to keep, no one is turned away. There is always a voice here. We help each person choose one with the right age, the right warmth, the right feel. A voice they are proud to be heard in.
We compare on price, on voice, and on dignity. The rest of the industry can keep the paperwork.
The promise
Your voice. The world listens.
Every word you speak comes out in your own voice: your banked voice when there is enough of it, or a voice you choose. CORA and Atlas have their own voices. They will never speak as you. Your voice stays yours. Always.
Who We Serve
Speech is lost in different ways, on different timelines, for different reasons. The system is designed to meet each one where it actually lives, never forcing one mold onto nine realities.
The body becomes a locked door. The person is still home.
Our answerTheir real voice, banked early. Voice, to lip reading, to eye gaze, on their timeline. Never forced.
Their page →The voice softens to a whisper while the person inside stays sharp and fully present.
Our answerSteady, forgiving targets designed to absorb the tremor. Their sound, banked before the whisper takes over.
Their page →Clear in the morning, gone by afternoon, back tomorrow. No two days the same.
Our answerBuilt to bend to the day they are having. High contrast for vision flares, bigger targets for fatigue, their banked voice for the rest.
Their page →The thought is complete. The path from mind to mouth is cut.
Our answerA distraction-free writing mode, word-by-word building, and a family taught how to wait, listen, and leave room.
Their page →The word is right there. It just will not line up when the room gets loud.
Our answerA quiet, single-focus mode and ready-made scripts for the high-stakes moments, waiting before the pressure hits.
Their page →The treatment that saves the life often takes the voice with it. Everyone knows the silence is coming.
Our answerThe voice is banked before surgery. Survival should not cost the sound of who you are.
Their page →The loss comes slowly, alongside a deeper one: being talked about instead of talked to.
Our answerLarge, calm tiles, a patient companion in CORA, and a voice quietly preserved as a gift the family keeps.
Their page →Language grows in whole scripts and phrases. Nearly every device fights the way it actually grows.
Our answerWhole phrases in a loved one's recorded voice, played exactly as recorded, then split and remixed as they are ready. The intonation is the meaning.
Their page →A breathing tube, a wired jaw, a healing throat. Fully alert, and unable to be understood at the moment they are most afraid.
Our answerSet up before the surgery, speaking for them in recovery, walked away from when the voice returns.
Their page →Before anything else
A diagnosis starts a clock on a person's voice, and nobody should wait on a device to start saving it. We bank voices at no charge, before you buy anything, whether or not you ever do. The recordings are theirs either way.
Start Free Voice BankingDesigned for the road ahead
Most devices are built for one frozen moment of ability. Grand Voice OS is designed to travel the whole road, and the order is the person's to choose.
FIRST
Speak naturally while CORA listens, and bank the real voice early, while it is easy. Every phrase saved is the person's own sound, kept.
WHEN NEEDED
When volume fades, the camera reads silently mouthed phrases and speaks them aloud, in the banked voice, not a synthetic stand-in.
ALWAYS THERE
When hands and voice both step back, the eyes carry the conversation: gaze plus blinks, on a keyboard laid out the way speech therapists already teach.
On their timeline. Never forced.
Grounded in research
Every design decision in Grand Voice OS is anchored to published research in augmentative communication, speech science, and human factors. A few of the foundations:
Clinical practice pioneered at Boston Children's Hospital showed that banking whole recorded phrases, with their real intonation, preserves identity in a way synthesized speech alone cannot. Bank Your Voice and Studio Script are designed around both: the banked voice and the banked phrase.
Costello, Boston Children's Hospital message banking practiceCommunication-partner training is one of the most cited unmet needs in AAC: fewer than half of ALS families ever receive guidance, and untrained partners are a leading driver of device abandonment. That is why every kit ships with How to Talk with Me, and why it speaks for itself.
Communication partner instruction literature, AAC abandonment studiesThe eye-gaze surface follows the Fitzgerald Key, the color-coded language layout used in speech therapy for decades. A returning AAC user is at home on day one instead of relearning someone's new idea.
Fitzgerald Key, Goossens & Crain modified color codingKeyword-based sentence composition research shows dramatic reductions in the effort a motor-impaired person spends per sentence. Grand Voice OS composes full sentences from a handful of gaze-selected keywords, and the person always chooses among candidates. Nothing is ever said for them.
Keyword-to-sentence AAC research, keystroke-savings literatureThe Natural Language Acquisition framework recognizes that many autistic communicators build language in whole gestalts: complete phrases whose exact wording and intonation carry the meaning. Studio Script plays recordings untouched, then splits and remixes them only when the person is ready.
Blanc, Natural Language Acquisition; recognized by ASHASpeech loss is rarely a single event. The published course of ALS, Parkinson's, and MS moves through stages, so the system is designed as a road: voice, then lip reading, then eye gaze, with the same phrases and the same banked voice traveling the whole way.
Condition progression literature across nine communitiesFull evidence dossiers, one per community, are live on each community's own page, with every finding cited in full.
The System
About a third of communication devices end up abandoned, and the research names the reasons: nobody around the person was equipped, and the device never fit them. So this was never built as a device. The person holds one surface. The people who love them hold the rest.
What the person holds
Every one of them works with the internet off. Voice, lip reading, eye gaze, word boards that move, live translation, CORA: all on the device, spoken in the person's own banked voice.
Everything below is built around this one screen. The device never has to reach out. The people around it reach in only as far as the person allows.
It speaks in airplane mode.
Everything in the box →
"I love you." Press the tile and the character performs the sign, in either art style. Every frame was checked against the real sign before it shipped.
Try it now
What everyone else holds
For the family and care team. Nothing here can interrupt the device or reach into it. Messages and boards wait outside, and the person picks them up when they choose.
The speech therapist, or the family
A speech-language pathologist (SLP) or a family member types a word. The studio draws the tile in the house style, cartoon or real life, and sends the finished board to the person's device. In the SLP Lounge, therapists find peers, support, and coaching, included as part of the person's team.
Answers the programming burden
The person and their care team
The same portal on a full screen, for longer sessions and setup work. The person decides who is in their circle, and who is not. SLP training on how to use the device, free of charge.
Answers nobody taught anyone
Family, wherever they are
The family and care team's window into the user's world. It only ever looks the way the user points it.
Answers the family kept on the outside
Anyone, before buying anything
For anyone who needs it, before buying anything and whether or not they ever do. A diagnosis starts a clock on a voice, and the clock does not wait for a device.
Answers the stranger's voice
Start free voice bankingOffline by design. Online by desire.
The kit
No app-store hunting, no setup maze, no extra purchases discovered at midnight. The kit arrives assembled around one person and speaks within minutes of opening.
Designed to work with no internet connection at all. Nothing to subscribe to before it speaks.
A diagnosis starts a clock on a person's voice. We bank it before it is gone. Keep it: the family always receives their own copy, because the voice is their property. Become it: the system speaks as them, in their own banked sound, improving over time.
Everything runs on the device itself. No cloud, no audio leaving the room. Privacy is not a setting here. It is the construction.
Categories and colors follow the Fitzgerald Key, the layout speech therapists have taught for decades, so returning users feel at home on day one.
Speak and be understood across languages, on the device and in real time, so a caregiver in another tongue still hears you, and you still hear them.
Every kit includes How to Talk with Me, a guide that teaches the family how to hold a conversation with someone using the system. Listening is half the gift.
The Say It menu, as it ships on every device.
Meet CORA™
CORA is the companion inside Grand Voice OS: she walks the first-time setup, coaches voice banking, keeps the long afternoons company, and listens any hour of the night. She runs entirely on the device, so the conversation never leaves the room.
Honest failure beats fake success. When CORA does not catch something, she says so and asks again. She is designed to never guess on someone's behalf, and never to put words in their mouth. For a person whose words are their lifeline, false recognition is betrayal. That single refusal shapes the whole product.
CORA speaks in her own voice. Your words are spoken in yours: your banked voice when you have it, a voice you chose when you do not.
“Good morning. How can I support your voice today?”
CORA, every morning, in her own voice. She wears the color of the person she serves.
Not a chatbot, a presence. She suggests, never decides, and tells the truth when she is not sure instead of guessing on your behalf. She learns your way over time.
Her own voice, never yoursThe assistant behind the scenes: the calendar, on-device search, the weather, appointment notes. He does the fetching so CORA can stay the relationship.
His own voice, never yoursPrivate by architecture.
The words a person fights hardest to say should be the safest words in the world. By design, Grand Voice OS processes everything on the device itself and is designed to keep working with no internet connection at all. Your account manages your kit, your voices, and your care plan. It never touches your conversations: no audio streaming to a server, no company listening in. Privacy here is not a policy. It is the architecture.
One honest footnote: the user can choose to switch on two connected features. Live search, and location sharing with the family members they pick. Both arrive off. Both stay the user's call, always.
Not cloud-dependent. Not data-hungry. Fully edge-computed. Fully private. Fully human.
The Investment
Legacy devices run $14,000 or more, before the months of paperwork. We priced Grand Voice OS so a family can decide at the kitchen table, not in an appeals process.
$6,997
Complete kit · one time
$127/month
Care plan
The old math: $14,000 for a stranger's voice. The new math: a complete system, in their own voice, for about half.
Grand Voice OS is a personal voice companion system, not a medical device, and no prescription is required. We do not process insurance, Medicare, or Medicaid. Upon request, we provide a structured Compliance and Functional Capability Packet for families to submit to their private insurer or HSA/FSA provider for independent reimbursement consideration.
Request Your System
Every kit is assembled around one specific person: their condition, their language, their voice. Tell us who we are building for and we will reach out within two business days.
Because each system is prepared for one person and onboarding is deliberately controlled, allocations are limited and granted first applied, first allocated. Early access is reviewed in waves based on production capacity, support readiness, and user fit.
Your voice still matters. And with Grand Voice OS, the world listens.
Request Your System