Cognitia AI
Your portable memory graph
The system of record for personal AI context — entities, temporal facts, behavioral signals, multimodal identity (faces, voices, images), and cross-source provenance in one unified graph.
Powered by MAGMA: a bi-temporal memory architecture where facts supersede on contradiction, every property traces back to its source, and you can replay the graph as it existed at any moment in time.
You own every byte
Full portability. Permanent deletion. Zero copies retained.
Bi-temporal ledger
Facts supersede on contradiction. Full provenance for every change.
Cross-ecosystem ingestion
Google, Microsoft, Apple, finance, social, and messaging as first-class sources.
Watch a memory graph form and self-correct
This is a fictional persona, Maya Chen, built from email, calendar, social, health, and chat. Press play on the timeline to watch two years of memory assemble itself, supersede stale facts (a job change, a move from San Francisco to Austin), and stay correct. Click any node to see exactly what is known and where it came from — including the photos, voice clips, and AI-generated images tied to each person.
Loading interactive demo...
MAGMA memory primitives
Six organized layers that map to the underlying graph — entities, facts, signals, and timeline data.
Entities & background
Canonical people, organizations, and long-term facts about who you are — deduplicated across sources.
Temporal facts
Bi-temporal state facts with supersession — when something changes, the graph updates correctly.
Events & timeline
Time-stamped records from email, calendar, transactions, and imported conversations.
Conversation memory
Compressed interaction history including imports from WhatsApp, Messenger, Instagram, TikTok, WeChat, and LINE.
Behavioral signals
Preferences and patterns with confidence scoring — how you communicate, what you prioritize.
Synthesized insights
Daily consolidation refines the profile — emerging themes and evolving preferences over time.
Multimodal memory
Faces, voices, and images linked to the people they belong to — per-person galleries, photo identify, and cloned-voice playback, portable over MCP.
How the graph builds
Connect sources — email, calendar, files, finance, social, messaging, local device
Extract entities, facts, and behavioral signals automatically from ingested data
Supersede stale facts as new evidence arrives — bi-temporal state tracking
Retrieve relevant memory for any authorized client — chat, Explorer, or MCP
Trace every fact to its source with full provenance
Delete per-fact, per-source, or export the entire graph
What Memory Looks Like
Cognitia transforms your interactions into structured memory: summaries, relevant facts, and temporal context for better decisions.
Example memory context block
<USER_SUMMARY> You are preparing a Tokyo trip in April and prefer nonstop flights when prices are comparable. You prefer calm neighborhoods, vegetarian-friendly restaurants, and morning schedules. </USER_SUMMARY> <RELEVANT_FACTS> - Prefers nonstop flights for long-haul travel. (2026-02-10 - present) - Budget target for flights is under $1,200 round trip. (2026-02-11 - present) - Wants 2 hotel options in Shibuya and Asakusa. (2026-02-12 - present) - Previously preferred evening meetings, now prefers morning meetings. (2025-11-01 - 2026-01-15) - Updated preference to morning meetings. (2026-01-16 - present) </RELEVANT_FACTS>
How you interact with your memory
Built-in chat and Memory Explorer help you interact with and inspect your memory. They are not the product — the graph is.