Multitenancy

Learn how to organize and manage data for multiple tenants in SourceSync.

Overview

SourceSync supports multitenancy through three primary mechanisms:

1. Native Tenant Support: Using X-Tenant-ID header for automatic tenant isolation
2. Namespaces: For complete data isolation between tenants
3. Metadata: For flexible tenant-specific organization within namespaces

Choosing Your Approach

Native Tenant Support (Recommended)

Best for organizations that need:

• Automatic tenant isolation without infrastructure changes
• Simple tenant management with header-based control
• Flexible scaling across tenants
• Cost-effective multi-tenant deployments

Implementation:

1. Add the X-Tenant-ID header to your API requests:

   curl -X POST https://api.sourcesync.ai/v1/ingest/text \
     -H "Authorization: Bearer $RAGAAS_API_KEY" \
     -H "X-Tenant-ID: tenant-123" \
     -H "Content-Type: application/json" \
     -d '{
       "namespaceId": "shared-namespace",
       "ingestConfig": {
         "source": "TEXT",
         "config": {
           "text": "Document content..."
         }
       }
     }'
   

   CopyCopied!

2. Documents are automatically isolated by tenant:

   • Each document is tagged with the tenant ID
   • Search and retrieval automatically scoped to tenant
   • No cross-tenant data access possible

3. Benefits:

   • No need for separate infrastructure per tenant
   • Automatic tenant isolation in shared namespaces
   • Simplified tenant management
   • Cost-effective scaling

Namespace-Based Isolation

Best for organizations that need:

• Complete data isolation between tenants
• Separate infrastructure per tenant
• Independent scaling and billing
• Compliance with data residency

Not suitable when:

• Content sharing between tenants is required
• Cost optimization is priority
• Managing many small tenants

Metadata-Based Organization

Best for organizations that need:

• Shared infrastructure across tenants
• Flexible content sharing
• Cost-effective solution
• Simple tenant management

Not suitable when:

• Strict data isolation is required
• Compliance requires separate infrastructure

Implementation

Using Namespaces

Create a dedicated namespace for each tenant:

curl -X POST https://api.sourcesync.ai/v1/namespaces \
  -H "Authorization: Bearer $RAGAAS_API_KEY" \
  -H "Accept: application/json" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "tenant1-production",
    "fileStorageConfig": {
      "provider": "S3",
      "bucket": "tenant1-docs",
      "region": "us-east-1",
      "credentials": {
        "accessKeyId": "...",
        "secretAccessKey": "..."
      }
    },
    "vectorStorageConfig": {
      "provider": "PINECONE",
      "apiKey": "...",
      "indexHost": "tenant1.pinecone.io"
    },
    "embeddingModelConfig": {
      "provider": "OPENAI",
      "model": "text-embedding-3-small",
      "apiKey": "..."
    }
  }'

Using Metadata

Add tenant information during document ingestion:

curl -X POST https://api.sourcesync.ai/v1/ingest/text \
  -H "Authorization: Bearer $RAGAAS_API_KEY" \
  -H "Accept: application/json" \
  -H "Content-Type: application/json" \
  -d '{
    "namespaceId": "shared-namespace",
    "ingestConfig": {
      "source": "TEXT",
      "config": {
        "text": "Company policy document...",
        "metadata": {
          "tenantId": "tenant-123",
          "department": "hr",
          "accessLevel": "internal",
          "region": "us-west"
        }
      }
    }
  }'

Filter content by tenant metadata:

curl -X POST https://api.sourcesync.ai/v1/search \
  -H "Authorization: Bearer $RAGAAS_API_KEY" \
  -H "Accept: application/json" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "company policy",
    "namespaceId": "shared-namespace",
    "filter": {
      "metadata": {
        "tenantId": "tenant-123",
        "accessLevel": "internal"
      }
    }
  }'

Examples

Document Search with Tenant Isolation

Search documents within a tenant's scope:

curl -X POST https://api.sourcesync.ai/v1/search \
  -H "Authorization: Bearer $RAGAAS_API_KEY" \
  -H "X-Tenant-ID: tenant-123" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "company policies",
    "namespaceId": "shared-namespace",
    "topK": 10,
    "filter": {
      "metadata": {
        "status": "published"
      }
    }
  }'

Document Management with Tenant Isolation

List documents for a specific tenant:

curl -X POST https://api.sourcesync.ai/v1/documents \
  -H "Authorization: Bearer $RAGAAS_API_KEY" \
  -H "X-Tenant-ID: tenant-123" \
  -H "Content-Type: application/json" \
  -d '{
    "namespaceId": "shared-namespace",
    "filterConfig": {
      "documentTypes": ["PDF", "DOCX"],
      "metadata": {
        "status": "published"
      }
    },
    "pagination": {
      "pageSize": 10
    }
  }'

Best Practices

1. Organization

   • Use clear naming conventions (e.g., {tenant}-{environment})
   • Define consistent metadata schema
   • Separate production and non-production environments

2. Security

   • Implement tenant validation in your application
   • Use metadata for access control
   • Consider regulatory requirements

3. Performance

   • Monitor resource usage per namespace
   • Use appropriate metadata indexing
   • Consider tenant size when choosing approach

Tenant ID Format

1. Use consistent, URL-safe tenant identifiers:

   • Alphanumeric characters
   • Hyphens and underscores
   • Lowercase recommended
   • Example: tenant-123, org-456-dept-789

2. Keep tenant IDs short but meaningful:

   • Include organization identifier
   • Add sub-organization if needed