Integrations
Connect your entire AI stack
Gateco plugs into the vector databases and identity providers you already use. No migration required. Add permission-aware retrieval on top of your existing infrastructure.
12
Vector Databases
4
Identity Providers
5 min
Avg Setup Time
Tier 1: Full Ingestion
RecommendedFull document ingestion, retroactive registration, and policy-aware retrieval. These connectors support the complete Gateco pipeline.
pgvector
GAPostgreSQL-native vector search with full SQL ecosystem support
Supabase
GAManaged Postgres with built-in auth, realtime, and vector support
Neon
GAServerless Postgres with branching, autoscaling, and pgvector
Pinecone
GAPurpose-built vector database with managed infrastructure
Qdrant
GAHigh-performance vector search engine with rich filtering
Tier 2: Retroactive Registration
Policy-aware retrieval with retroactive vector registration. Connect existing vector stores and gate access without re-ingesting data.
Milvus
PreviewScalable open-source vector database for billion-scale search
Chroma
PreviewDeveloper-friendly embedding database for AI applications
OpenSearch
BetaDistributed search with k-NN vector support and analytics
Azure AI Search
BetaMicrosoft Azure fully managed search with vector, keyword, and hybrid retrieval
Vertex AI Vector Search
PreviewGoogle Cloud managed vector database with ANN nearest-neighbor search for low-latency RAG retrieval
Vertex AI Search
PreviewGoogle Cloud managed enterprise search with native keyword, vector, and hybrid retrieval for RAG pipelines
Search mode coverage
Gateco supports 4 search modes: vector, keyword, hybrid, and grep. Coverage varies by connector. The table below shows what each connector actually supports today.
| Connector | Vector | Keyword (ranked) | Hybrid | Grep |
|---|---|---|---|---|
| PostgreSQL (pgvector) | ✓ | ✓ | native | regex |
| Supabase | ✓ | ✓ | native | regex |
| Neon | ✓ | ✓ | native | regex |
| Pinecone | ✓ | — | — | — |
| Qdrant | ✓ | filter only | native | substring |
| Weaviate | ✓ | ✓ | native | substring |
| Milvus | ✓ | filter only | native | substring |
| Chroma | ✓ | — | — | regex |
| OpenSearch | ✓ | ✓ | native | regex |
| Azure AI Search | ✓ | ✓ | native | — |
| Vertex AI Vector Search | ✓ | — | — | — |
| Vertex AI Search | ✓ | ✓ | native | — |
Pinecone: vector only
No keyword, hybrid, or grep support. Use Pinecone when you need pure ANN retrieval and manage re-ranking in your application layer.
Vertex AI Vector Search: vector only, no text content
No keyword, hybrid, or grep support. The connector never returns raw text — content is always resolved via sidecar metadata. Plan ingestion accordingly.
Chroma: grep only
No ranked keyword search or hybrid support. Chroma matches via substring or regex patterns only. Scores are uniform (1.0) — not ranked by relevance.
Maturity labels: GA connectors are fully supported and production-ready. Beta connectors are stable for retrieval but may have limited edge-case coverage. Preview connectors are available for evaluation with known limitations documented above.
Identity Providers
Growth+Sync your existing user directories for identity-aware retrieval policies. Map organizational roles and groups directly to data access rules.
Azure Entra ID
Microsoft identity platform with MS Graph API integration
AWS IAM
AWS Identity Center with Identity Store API integration
GCP Cloud Identity
Google Cloud identity with Admin SDK integration
Don't see your stack?
We're adding new connectors regularly. Tell us what you need.