🚀 EDDI 6.5.0

19 LLM providers, knowledge bases that crawl your site and read your files, and conversations that follow compatible agent versionsWhat's new

Managed Cloud Get Started
All integrations
Data and RAGstoreType: pgvector

pgvector

PostgreSQL with pgvector as the vector store for an EDDI knowledge base. The recommended persistent choice when you already run PostgreSQL.

ConfigurationDocs ↗

What you get

  • A knowledge base inside a database you already back up and operate
  • Verified document replacement on re-ingestion
  • The same agent configuration as every other store, with only the store parameters changing

How it works in EDDI

  1. Enable the pgvector extension in PostgreSQL and create a database and user.
  2. Create a RagConfiguration with storeType: "pgvector", the connection details, a table and the vector dimension.
  3. Bind it with an eddi://ai.labs.rag workflow step.
  4. Select it by name in the LLM task and ingest documents.

Configuration

Taken from the EDDI documentation. Placeholders such as your-agent-id are yours to replace.

RagConfiguration (POST /ragstore/rags)The dimension must match your embedding model. 1536 matches text-embedding-3-small.
{
  "name": "product-docs",
  "embeddingProvider": "openai",
  "embeddingParameters": {
    "model": "text-embedding-3-small",
    "apiKey": "${vault:openai-key}"
  },
  "storeType": "pgvector",
  "storeParameters": {
    "host": "localhost",
    "port": 5432,
    "database": "eddi",
    "user": "eddi",
    "password": "${vault:pg-password}",
    "table": "product_docs",
    "dimension": 1536
  },
  "chunkStrategy": "recursive",
  "chunkSize": 512,
  "chunkOverlap": 64,
  "maxResults": 5,
  "minScore": 0.6
}
workflow step + langchain.json taskBinding the knowledge base to the agent and using it in a task.
// 1. workflow step: binds the knowledge base to the agent
{ "type": "eddi://ai.labs.rag",
  "extensions": {},
  "config": { "uri": "eddi://ai.labs.rag/ragstore/rags/{ragId}?version=1" } }

// 2. LLM task: selects it by the "name" of the RagConfiguration
{ "tasks": [{
    "actions": ["*"],
    "type": "openai",
    "knowledgeBases": [{ "name": "product-docs", "maxResults": 5, "minScore": 0.7 }],
    "parameters": { "systemMessage": "You are a helpful assistant." }
} ] }

Governance

  • The database password is a required parameter. Keep it as a vault reference.
  • Your existing PostgreSQL access control and backups apply to the vectors.
  • Embedding calls send document text to the embedding provider you configure.

Good to know

  • The in-memory store is for development: it empties after 30 minutes without a query and on restart. A persistent store such as pgvector is the production choice.
Get startedDocs ↗

Related integrations

Data and RAGChromaChroma as the vector store for an EDDI knowledge base. The lowest-friction option when you have no existing infrastructure to reuse.Data and RAGElasticsearchElasticsearch as the vector store for an EDDI knowledge base, useful when you already run it and want keyword and vector search in one place.Data and RAGMongoDB Atlas Vector SearchMongoDB Atlas Vector Search as the vector store for an EDDI knowledge base, a natural fit if you already run on Atlas.

Configuration and behavior on this page follow the EDDI documentation, checked on 2026-10-14. Field names and defaults can change between releases, so confirm them in the docs for your version.