Parcle Personal Brain
API

Memory Events

Your Personal Brain keeps memory in layers: the raw chats and files you bring in, semantic memory events on a timeline, and the relationships between them.

Memory is where everything you bring in lives. It is much more than storage. Raw retrieval is only the floor. On top of it, the brain distills your chats and files into events, learns how they relate, and keeps them on a timeline. The result is a memory that understands your history, not just one that stores it.

A layered memory#

Raw Store Your chats and files · vector embeddings · keyword index Memory Events Each event = a description + a timeline The primary unit of memory: raw source opened only on demand Connected Memory Relationships between your events distill relate
Memory is layered. A raw store keeps your original chats and files, their embeddings, and a keyword index so anything can be found. Above it, the brain distills your data into events, each with a description and a timeline, and maps the relationships between them. The raw source is opened only when an event needs more detail.

The raw store#

At the bottom, the brain keeps your original chats and files alongside vector embeddings and a keyword index, so anything can be found either by meaning or by exact term. This makes your source material fully retrievable, but on its own it is a flat pile of text. It is the foundation, not the substance.

The semantic layer: memory events#

The substance of memory lives one layer up. The brain analyzes the raw material and distills it into events: meaningful things that happened, extracted from your data. Each event carries:

  • A description: what happened, in concise, structured form.
  • A timeline: when it happened and how it unfolded.

Events are the primary unit of memory. When an agent asks something, the brain reasons over events first and only opens the original chat or file when an event needs more detail. It then maps the relationships between events, building a connected memory rather than a disconnected archive. This is what lets your brain answer with context and continuity instead of isolated snippets.

Merged timeline of chats, files, and activity tracked over time.
The Personal Brain keeps a merged timeline of your activity, so agents can retrieve not just what happened, but when it happened, what changed, and which chat or file it came from.

Three kinds of memory#

The brain draws on three distinct kinds of memory, each with a different source, reliability, and lifecycle:

Factual MemoryExperiential MemoryWorking Memory
SourceRead directly from your chats and filesInferred and learned from how you workThe current chat, your instructions, and task context
CorrectnessPlainly true; changes only when your data changesVerified and refined over time; when unclear, the agent can ask youSession-bound while active; useful records are saved back into persistent memory
LifecyclePersistent; re-synced when you add or change dataPersistent; continuously evolvingUsed during the task, then promoted into memory with what's worth keeping

Private by default#

All of this is yours alone. Your memory is scoped to your user id and never shared with another user's brain. Ingestion, search, and deletion all operate inside that boundary. Delete a chat, a file, or anything under a tag, and it is gone from your memory. See the Memory API for how ingestion, search, and deletion work.

🗂️Next: Ontology Graph

How your memory becomes a semantic map of the concepts you use.

Parcle Personal Brain · Documentation