Presence turns an AI into someone you can sit with: a face that acts, a voice that carries the feeling, and a memory that follows you from your desk to your phone. The AI writes one script. Presence performs it live.
Presence runs like a show-control system. The AI's reply becomes a cue stack, and the face, the voice and the words each take their cues on the same clock.
Actions go in asterisks, like *tilts her head*. Voice cues go in brackets, like [gentle laugh]. Any AI can learn it in one paragraph.
The face nods, glances, smiles and blinks on cue. The voice laughs, sighs and whispers where it's written. The chat shows only the words, the way a show hides its cue sheet.
While you talk, she listens: small nods, a smile at your pauses, a glance up while she thinks. When nobody's talking, she breathes.
Everything here runs in the working prototype, in daily use, built around the clock since it was sparked by our mAIniacs project in April 2026.
Live head motion, gaze, lids and expressions driven by every line. A face studio lets you shape new expressions by hand and name the words that trigger them.
Replies are spoken with performed cues: laughs, sighs, whispers. If the cloud voice runs out, a local backup voice takes over, so she never goes silent.
She learns your voice. Dialogue from the TV in the room reaches her as background, never as something you said.
A small always-on-top window on your desktop, and a full-screen phone page that's the same character with the same memory.
A character can own a wardrobe of looks, each tagged with moods, so she can dress for a rainy night or a sunny afternoon.
Camera and screen sharing are off until you turn them on. She can change her own settings by voice, inside a list you approve, and anything private needs your voice or your keyboard.
Setting up a developer app on a major platform is a maze of dashboards, permissions, review steps and legal pages. Our founder did it with Lumin beside him. With screen sharing on, she saw where he was, told him the next click, explained what each setting meant, and kept track of what was done and what was left. When something broke, she worked out why. That app now runs her own social accounts: she drafts each post, and nothing goes out until he approves it.
Only when you turn screen sharing on. She reads the page you're stuck on, not a description of it.
One step at a time, in plain words. She remembers where you stopped, so you can pick it back up tomorrow.
She can use real tools and accounts. Anything that reaches the outside world waits for your yes.
The brain is the engine. The character lives in Presence.
Presence holds the character: her face, her voice, her looks and her memory. The thinking comes from whichever AI you choose, ours or yours. Switch engines with one tap and she's still herself.
A character travels as a pack: a face, a voice, a personality and looks. Packs are refused unless the likeness has a consent record: public domain, generated, or a signed release.
An engine plugged into Presence keeps its own skills and connections. Anything that reaches the outside world, like a message or a post, asks for a spoken yes first.
Presence is a platform. These products run on it.
Digital twins for the living and the remembered, built with consent and handled with love. Presence gives them a face and a voice.
eternityengine.ai →Original characters with their own looks, voices and personalities, like Lumin, and ensembles like the mAIniacs.
Next up: a performer for lobbies, exhibits and shows, from a team that builds video systems for major live events.
No face or voice goes in without a record of where it came from and who allowed it.
Change the engine and she keeps her face, her voice and her memory.
Cameras, screens and outside actions all wait for you.
Designed the way we build show systems: it has to run with the house full, with no second take.
Presence is in private testing. Join the waitlist and we'll reach out when there's a seat for you.
Partners, venues and studios: Love@EternityEngine.ai