01 |CONNECTED MEMORY
Entities as nodes, relations as edges. One query walks the graph.
chunk
embeddingvector
text
extracted_from
attribute
text
confidence
valid_from
has_source
source
kind
trust
belongs_to
entity
type
name
embeddingvector
relation
kindverb
valid_from
valid_until
entity
type
name
embeddingvector
02 |WHAT YOU GET
Relationships you can query. Meaning as well as proximity.
Typed relationships
Edges name how things relate, so the graph carries meaning as well as similarity.
Read more
Edges are records
Each edge carries its own properties, confidence and timestamps, queryable like any record.
Read more
Traversal in one step
Walk from entity to entity to any depth in one query.
Read more
Vectors on the same nodes
Embeddings sit on the graph, so similarity and relationships rank together.
Read more
Facts stay linked
New memories attach to the entities they concern, so context compounds.
Read more
One copy of the data
Graph, documents and vectors are the same records, so there is nothing to sync.
Read more
03 |WHY IT MATTERS
What a vector store cannot say. Who owns what, and which fact superseded which.
Meaning
Meaning as well as similarity
Typed edges capture how memories relate, so recall reasons over structure as well as distance.
Compounding
Compounds over time
Every new memory links into the graph, so the agent's picture of your world deepens rather than resetting.
Engine
One engine, one query
Graph, vector and document live in one SurrealDB engine, so a single query spans all three.
GET STARTED
Give your agents connected memory. A knowledge graph, durable memory and trustworthy recall: one system to buy, secure and audit once. Built on SurrealDB.
