User memory is AGNT's per-user fact store: key-value pairs with confidence scores and pgvector embeddings, split into structural keys that always enter the prompt and semantic keys recalled on demand.
Last verified: April 2026
What is User Memory?
User memory is AGNT's per-user fact store: key-value pairs with confidence scores and pgvector embeddings, split into structural keys that always enter the prompt and semantic keys recalled on demand.
Each fact row carries a key, a value, a confidence between zero and one, a timestamp, and a 1536-dimensional embedding. Structural keys (diet, interests, birthday, fitness goal, preferred booking hour) are always injected into the system prompt. Semantic keys are recalled per conversation through cosine distance search against pgvector.
FAQ
User Memory FAQ.
Common questions about User Memory in the AGNT platform.
User memory is AGNT's per-user fact store: key-value pairs with confidence scores and pgvector embeddings, split into structural keys that always enter the prompt and semantic keys recalled on demand.
People also ask.
Related terms
Structural Keys
Structural keys are the user-memory fields that are always included in the system prompt — diet, interests, last_booking, typical_party_size, fitness_goal, and a dozen more — so they never require embedding search.
Semantic Recall
Semantic recall is the embedding-based retrieval step that finds memories relevant to the current message, even when the wording has changed since the original fact was stored.
Soul Loader
The Soul Loader is AGNT's system-prompt builder: it assembles the active prompt from user memory, venue soul, time-of-day context, and recent activity on every turn.
Personal Agent
A Personal Agent is an individual user's AI, living inside WhatsApp, Telegram, or Instagram, that handles venue search, bookings, calorie scanning, dupe search, and transport on their behalf.
Source: agnt-backend/app/db/models.py
Share as social post
What is User Memory? User memory is AGNT's per-user fact store: key-value pairs with confidence scores and pgvector embeddings, split into structural keys that always enter the prompt and semantic keys recalled on demand. https://agntdot.com/glossary/user-memory
263 / 280 chars
See it in action.
Now you know what User Memory means. Try the live scan demo or read the developer docs to go deeper.