peepshow/ sinks/ vector-memory

Category ◆ 14 sinks ◆

peepshow sink category

Vector + AI memory

Index frames for semantic recall — Chroma, Qdrant, Pinecone, Milvus, MemPalace.

Vector DB

Chroma

Index frames into a Chroma vector DB collection.

  • search
  • memory
  • llm-pipeline
Vector DB

Qdrant

Upsert frame points into a Qdrant collection.

  • search
  • memory
  • llm-pipeline
Vector DB

Pinecone

Upsert frames into a Pinecone serverless index.

  • search
  • memory
  • llm-pipeline
Vector-in-SQL

pgvector

Vector search on top of the postgres sink.

  • search
  • memory
  • llm-pipeline
Local AI memory

MemPalace

Mineable markdown notes into a local MemPalace.

  • memory
  • llm-pipeline
Agent memory

Zep

Append every run to a Zep long-term-memory session.

  • memory
  • llm-pipeline
Agent memory

Mem0

Distil every run into Mem0 long-term memory.

  • memory
  • llm-pipeline
Vector DB

Weaviate

Batch-insert frames into a Weaviate class.

  • search
  • llm-pipeline
Vector DB

Milvus

Upsert frame rows into Milvus or Zilliz Cloud.

  • search
  • llm-pipeline
Agent memory

Letta

Post a run summary to a Letta (MemGPT) agent inbox.

  • memory
  • llm-pipeline
Full-text + search

Elasticsearch

Index each run as a doc — full-text search across run history.

  • search
  • archive
  • memory
Hybrid search

Meilisearch

Self-hosted hybrid search — POST runs at a Meilisearch index.

  • search
  • archive
  • memory
Hybrid search

Typesense

Fast hybrid search — upsert runs into a Typesense collection.

  • search
  • archive
  • memory
In-memory search

Redis Vector

Hash + RediSearch index — every run becomes a Redis 8 hash.

  • search
  • memory
  • llm-pipeline

← All peepshow sinks

In Vector + AI memory

Other categories