Connect Neon to Gateco

Neon is a serverless Postgres platform with built-in pgvector support. Gateco treats Neon as a Tier 1 connector — supporting full document ingestion, retroactive registration, and RAG access control. This guide covers the complete setup from a fresh Neon project to your first identity-aware retrieval, adding vector database RBAC on top of your existing pgvector tables.

Step 1 — Get your Neon connection string

1. Log in to console.neon.tech and open your project.
2. In the Connection Details panel, select your database and role.
3. Copy the connection string. It looks like:

postgresql://user:password@ep-cool-term-123456.us-east-2.aws.neon.tech/dbname?sslmode=require

Step 2 — Enable the pgvector extension

Connect to your Neon database with any SQL client and run:

CREATE EXTENSION IF NOT EXISTS vector;

Verify it is installed:

SELECT * FROM pg_extension WHERE extname = 'vector';

Step 3 — Create a dedicated database user (recommended)

Using a dedicated user makes it easy to audit Gateco access and rotate credentials independently.

-- Create the Gateco user
CREATE USER gateco WITH PASSWORD 'your-strong-password';

-- For ingestion (write access to your vector table)
GRANT INSERT, SELECT, UPDATE, DELETE ON your_vector_table TO gateco;
GRANT USAGE, SELECT ON SEQUENCE your_vector_table_id_seq TO gateco;

-- pgvector catalog access (required for metadata introspection)
GRANT SELECT ON pg_extension TO gateco;

Step 4 — Add the connector in Gateco

1. In the Gateco dashboard, navigate to Connectors → Add connector → Neon.
2. Paste your connection string from Step 1.
3. Click Test connection. Gateco will introspect your schema and show available tables.
4. Click Save.

Step 5 — Configure search settings

After saving the connector, configure the search settings to point Gateco at your vector table:

1. In Connectors, click your Neon connector → Search config.
2. Set the Vector table name (e.g. embeddings).
3. Set the Embedding column (e.g. embedding).
4. Set the Content column if your table stores the original text (e.g. content).
5. Save the search config.

Search configuration reference

Ingest your first document

from gateco_sdk import GatecoClient

client = GatecoClient(api_key="gck_live_...")

# Ingest a document
result = client.ingest.document(
    connector_id="your-connector-id",
    resource_id="doc-001",
    content="This document contains confidential HR information.",
    embedding=[0.1, 0.2, ...],  # your 1536-dim embedding
    metadata={"classification": "confidential", "department": "hr"},
)

Troubleshooting