Stage 03 · Research programme
SoulHub Avatar
Persistent presence, any surface
The same presence, everywhere you already are.
SoulHub Avatar · Research programme In one paragraph
SoulHub Avatar — SoulHub Avatar lifts the presence off the device. The same memory vault drives a persistent avatar that can appear in spatial computing environments, on a phone, or in a shared virtual space — so continuity of memory is not tied to a single object in one room.
Capabilities
What it does differently
One memory, many surfaces
The vault is the product. Avatar simply renders it wherever you are, without forking the memory into copies.
Spatial conversation
Avatars occupy a position in a room, hold eye contact and take turns, rather than floating as a video tile.
Shared family spaces
Relatives in different countries can sit in the same space with the same presence at the same time.
Consent travels with it
The permissions recorded when a vault is created follow the avatar to every surface it appears on.
Specification
At a glance
| Form | Software avatar, device-independent |
|---|---|
| Targets | Spatial headsets, mobile, shared spaces |
| Continuity | One vault across every surface |
| Model | Memory Graph v2, streamed |
| Governance | Consent record travels with the avatar |
| Status | Research programme |
Inside
Every SoulHub Avatar runs Alaya AI.
The display is the visible half. The other half is a twelve-stage cognitive pipeline — memory retrieval, identity assembly, affect shaping, two verification passes and four guardrails — that runs on every turn.
PerceptionĀlaya retrieveManas identityCaitta affectCognition generateSelf verificationWitness verificationConsistency guardTrisvabhāva guardWellbeing guardCaitta expressĀlaya perfume
Inside Alaya AI → 