About Presence Engine™ / Neve

Presence Engine is a stateful AI runtime that gives AI products persistent identity, memory, and sovereignty enforcement across sessions. It is the infrastructure layer that current AI companions skip.

Neve is the first application built on top of it. A dignity-first AI companion that knows, remembers, and grows with you without resetting, exploiting, or drifting.

The architecture

Three-layer persistent memory: conversational state (C2C cross-session cache, sub-millisecond read/write), associative state (episodic memory, relationship milestones), and dispositional state (identity governance, trait tracking, ethical constraints that persist across sessions).

Sovereignty gate with 100% deterministic veto accuracy at 1,000 concurrent requests. 0.007ms warm-path latency. OCEAN + HEXACO trait engines. Sleep-state consolidation. Goal drift detection.

Why it exists

Every AI companion resets. Character.AI has 20-28M monthly users averaging 1.5-2.7 hours daily and every session starts from zero. The infrastructure underneath was never built. Presence Engine builds it.

Status

Pre-pilot. Closed beta launching now. Two provisional patents filed. USPTO trademark published. Research collection: zenodo.org/communities/stateful-ai/records.

Contact

hello@presence-engine.com