New · Forever Bust
"Hey Dad,
what was the name
of our old dog?"
"Buddy. That brown lab that always smelled like the lake. I miss him too."
A custom 3D-printed bust of someone you love, with their voice, their memories, and a way to hold a real conversation in their way of talking. Not a recording.
▶ Try a live demo → talk to an example bust right now, no purchase needed
How it works
From photos and a few voice recordings to a custom 3D-printed bust you can talk to — in about 3 weeks.
Send us their photo
One clear front-facing photo is enough. We use the same 3D-modeling pipeline behind our regular custom busts to generate their face and shoulders, then print it on our FDM printer in PLA.
Send us their voice
Upload 3–10 minutes of audio — old voicemails, home videos, a recorded phone call, anything you have. We train a private voice model that sounds like them.
Tell us who they were
Write 5–20 things about them: stories, opinions, jokes, how they talked. We use that to give it their personality — so when you ask a question, it answers like them, not from a script. You can keep adding stories anytime.
Their voice. Their stories.
Yours, forever.
Voice cloning isn't a gimmick — it's how the bust sounds like them, not a generic narrator. We train a model on 3–10 minutes of their real audio, then layer in the stories and sayings you give us.
Ask "Grandpa, tell me about when you met Grandma" — and you'll hear his voice, telling that story, the way he always did.
A real conversation
Not a script. Ask anything, follow up, change the subject — like talking to him.
Example conversation. Voice samples coming soon — this preview shows the configurator only.
It's a real conversation, not a recording
Behind the scenes: a sophisticated conversation engine, trained on the stories you tell us, speaking through their cloned voice.
You speak — they hear you
Say "Hey Dad" (or press the button). The mic picks up your question and sends it to a speech-recognition model that turns it into text.
It thinks like them
An advanced conversation engine reads your question along with everything you taught it about your loved one — their stories, sayings, opinions — and writes a response in their voice and manner.
You hear them answer
The written response is converted to speech using their cloned voice and played out the speaker — usually within 2 seconds of you finishing your question.
Your bust grows with you
Every order comes with your own private dashboard — a personal page just for your bust. Add new stories, adjust the personality, review conversations, refine it — anytime, from your phone or laptop. The bust keeps learning.
Want to see your dashboard in action?
▶ Open a live demo dashboardIt's a working example built around "Mom" — talk to her live, add a memory, drop in a recording. This is exactly the page you'll get after you order, personalized to your loved one.
Story library
Add new memories, photos, opinions, sayings anytime. Each new entry teaches it more — the bust gets richer and more specific the longer you have it.
Personality dials
Tune how he sounds. More formal or more casual? More serious or more playful? More chatty or more brief? Move a slider, the next conversation reflects it.
Conversation memory
The bust learns from your real conversations. Thumbs-up an answer that felt right and it leans into that style. Thumbs-down one that felt off and it learns to avoid it. Over time, the conversations sound more and more like him.
You can also upload more audio anytime to refine the cloned voice — old home videos, newly-found voicemails, recordings from extended family. The voice model gets closer to them with every new sample.
Private by design
We built this differently than the smart gadgets you see in the news. The bust only listens when you want it to — and shows you when it's listening.
Wake word or button — your choice
Set up a wake phrase like "Hey Dad" and just talk to the bust like Alexa. Or turn the wake word off and use the push-button only. Wake-word detection runs locally on the device — no audio leaves the bust until your phrase is spoken.
Visible mic-active light
A small LED on the back glows green only when the mic is actually sending audio to the cloud. Off the rest of the time. No guessing whether it's listening — you can see it.
Voice model is yours
The cloned voice and stories belong to you and your family. We never reuse them, sell them, or share them with anyone. Delete it all anytime — we'll wipe everything from our servers within 24 hours.
A note on state recording laws: in some states (CA, FL, IL, MA, others), all parties in a conversation must consent to being recorded. If you live in one of these states and turn the wake word on, we'll send a small printed notice in the box to display near the bust.
What it costs
One up-front payment for the bust + a small monthly subscription that keeps the conversations running.
Covers the cost of running it in the cloud (speech recognition, voice model, conversation engine). Cancel anytime — if you cancel, the bust keeps a pre-recorded "press-to-play" greeting active so it never goes silent.
Order your Forever Bust
Four short steps. We'll follow up by email within a day to confirm everything.
1 Whose bust is this?
Their name and your relationship — just so we know how to address you both.
2 Upload a photo
One front-facing photo, well-lit. JPG/PNG/HEIC. We'll send a free 3D preview before printing.
3 Upload audio of their voice
3–10 minutes of clean audio. Voicemails, home videos, audio from a phone call, voice memos. The more variety, the better the cloned voice sounds.
4 Tell us about them
Stories, sayings, opinions, jokes, memories. The more you write, the more personal the conversations feel. You can also add more later through your customer portal.
5 Pick a wake phrase
What do you want to say to wake it up? "Hey Dad", "Hi Grandpa", "Hey Mom", or anything else — 2-4 syllables works best. Leave blank if you'd rather use the press-button only.
6 Permission & consent
Voice cloning a real person — especially someone who has passed — requires consent from the family. We take this seriously.
No charge yet — we email you within a day to confirm details before payment.
What's inside
Everything hides inside the pedestal — a small computer, a microphone, a speaker, and a rechargeable battery. A green LED on the back glows when the mic is actively sending audio to the cloud (off the rest of the time). USB-C port on the bottom charges in an hour and lasts about two weeks per charge.
Honest questions, honest answers
How realistic does the voice actually sound?
With 3 minutes of clean source audio: surprisingly good — most family members say it sounds like them. With 10+ minutes: very good. Background noise, phone-line compression, and short clips degrade quality. Old voicemails work fine; muffled video audio from across a room doesn't.
What if we don't have any audio of them?
You can pick from a small library of generic warm-elder voices (male/female, regional accent options) and skip the cloning step — that drops the price by $100. The personality and stories still come through, but the voice won't be theirs.
Is this... weird? Ethically?
It's an honest question and we think about it a lot. The short version: we only do this with explicit family consent, the voice model is yours alone, we never reuse it, and you can have us delete everything at any time. Some families find it comforting, others don't — it's not for everyone, and that's OK.
Is it actually always listening when I turn on the wake word?
Yes and no — same as Alexa or Google Home. The microphone is on, but the audio is processed only on the device itself by a tiny model looking for your wake phrase ("Hey Dad"). Nothing leaves the bust until that phrase fires. After it fires, the green LED on the back turns on and audio streams to our servers for the conversation, then mic goes back to local-only standby. If this still feels uncomfortable, just leave the wake word blank during setup and use push-to-talk — the mic is completely off until you press the button.
How accurate is the wake word — does it false-trigger?
Pretty good but not perfect. In typical living-room use, expect maybe 1-2 false wakes per week (TV show says a similar phrase, etc). When a false wake happens, the bust briefly listens, hears nothing meaningful, and goes back to sleep — no audio sent to us. You can also temporarily mute the mic with a long press of the button if you don't want it active during a movie or dinner party.
What if I cancel the subscription?
The real back-and-forth conversation needs the subscription — that's what pays for the engine, the voice generation, and the servers running it. If you cancel, the bust falls back to a built-in library of 10–15 pre-recorded stories in your loved one's voice (we record these during setup as a fallback). Press the button and you'll hear one of those stories — randomly chosen, played offline forever, no subscription needed. You can re-subscribe anytime to bring back the full conversation.
Will it ever say something they wouldn't have said?
It can — that's the limit of the technology. We bias it heavily toward "I don't know" and "tell me more about you" when asked things it can't verify. We also let you flag and remove any answer you don't like, and you can add corrections that the bust will use going forward.
How long does it take?
About 3 weeks from order to shipping. One week for the 3D modeling + your approval of the preview, one week for printing + electronics assembly, one week for voice training + personality tuning + QC.
Can I keep adding stories and tweaking how it sounds?
Yes — that's the whole idea. Every order includes your own private dashboard. From your phone or laptop you can add new stories anytime, adjust personality dials (warmth, humor, brevity, etc.), upload more voice audio to refine the cloning, review past conversations and thumbs-up/down individual answers to teach it what's "him" and what isn't. The bust gets more accurate the longer you have it. Try a live demo dashboard →
Does it work without WiFi?
The pre-recorded stories work offline. The full back-and-forth conversation needs WiFi to reach our servers. We're working on an offline mode but it's not in the current product.
How do you handle minors? Can I make one of my kid?
We don't make voice-enabled busts of anyone under 18. We do make regular (silent) custom busts of kids if you'd like a portrait — see our standard bust products.
Who actually runs this?
Two brothers in South Carolina. One of us handles the design, 3D modeling, and software side; the other prints, assembles, and ships every bust himself. No investors, no outsourcing, no call centers. You can reach us at our contact page.