Personal Replicant

World-Class Digital Twin Ecosystem

VRAM OVERHEAD

0.0 / 96 GB

Demo

Portfolio showcase running with realistic sample data. The full app runs locally against a Python + ChromaDB backend and dual RTX A6000 GPUs — here every panel is populated with a mock digital-twin persona so you can explore the complete UI.

Overview — The Memory Gatekeeper

This is the front door of the system. You write a raw memory — a fact, story, belief, or preference — and the AI classifies it into one of six identity pillars, extracts the people mentioned, and refines the wording before it is committed to the vector database. Everything the twin 'knows' starts here.

1Write a memory2Run AI analysis3Review pillar & people4Commit to memory bank

Memory Gatekeeper

Active Intelligence

Ingest New Memory

Input Stream

Identity Pillar

Impact Weight