Connect Qdrant to Gateco

Qdrant is a high-performance vector search engine with rich Qdrant payload filtering capabilities. Gateco connects to Qdrant as a Tier 1 connector, supporting full document ingestion, retroactive registration, and RAG access control across Qdrant collections — adding identity-aware retrieval and vector database RBAC without replacing your existing Qdrant collection access setup.

Qdrant keyword search uses payload text matching (not ranked BM25). All matched documents return uniform score=1.0. For ranked keyword search, consider using a Tier 1 Postgres connector.

Step 1 — Create a Qdrant collection

If you already have a collection, skip to Step 2.

bash

curl -X PUT 'http://localhost:6333/collections/my_collection' \
  -H 'Content-Type: application/json' \
  -d '{
    "vectors": {
      "size": 1536,
      "distance": "Cosine"
    }
  }'

Or via the Qdrant Console:

Log in to cloud.qdrant.io (or open your self-hosted dashboard at :6333).
Navigate to Collections → Create collection.
Set vector size (must match your embedding model) and distance metric.
Click Create.

Step 2 — Get your connection URL and API key

For Qdrant Cloud:

In the Qdrant Cloud console, click into your cluster.
Copy the cluster URL (e.g. https://abc123.eu-central.aws.cloud.qdrant.io:6333).
Go to API Keys → Create API key. Copy it — it is shown only once.

For self-hosted Qdrant, your URL is typically http://your-host:6333. No API key is required unless you have configured one.

Step 3 — Add the connector in Gateco

Navigate to Connectors → Add connector → Qdrant.
Enter your URL, collection name, and API key (leave blank for unauthenticated self-hosted).
Click Test connection.
Click Save.

Field	Example	Description
`url`	`https://abc123.cloud.qdrant.io:6333`	`Qdrant cluster or instance URL`
`collection_name`	`my_collection`	`Name of the Qdrant collection`
`api_key`	`qdrant-api-key-here`	`API key (Qdrant Cloud only; leave blank for self-hosted)`

Troubleshooting

Error	Cause	Fix
`Forbidden (403)`	`Wrong or missing API key`	`Verify the API key in Qdrant Cloud → cluster → API Keys`
`Collection not found`	`Wrong collection name`	`Check the exact collection name in the Qdrant console (case-sensitive)`
`connection refused`	`Qdrant not reachable`	`Verify the URL, port (default 6333), and that your firewall allows traffic from Gateco`

Frequently asked questions

Does Gateco work with both Qdrant Cloud and self-hosted Qdrant?

Yes. For Qdrant Cloud, enter the cluster URL (e.g. https://abc123.eu-central.aws.cloud.qdrant.io:6333) and the API key from the cluster settings. For self-hosted Qdrant, use your host URL (e.g. http://localhost:6333). Self-hosted instances do not require an API key by default — leave the API key field empty.

Is an API key required for Qdrant?

An API key is required for Qdrant Cloud. For self-hosted Qdrant, API keys are optional — they can be configured via the service.api_key setting in qdrant.yaml. If your self-hosted instance has no API key configured, leave the field empty in the Gateco connector settings.

Does Gateco expect a specific collection structure?

No specific payload structure is required for policy-aware RAG retrieval. Gateco stores its access control metadata in a sidecar store separate from the Qdrant payload (the default metadata_resolution_mode: sidecar). If you want Gateco to read policy metadata directly from the Qdrant payload — useful when Qdrant payload fields already encode ownership or classification via Qdrant payload filtering — switch to metadata_resolution_mode: inline and configure metadata_field_mapping in the connector's search config.