Bring Your Own Storage (BYOS) Instead of Mandatory IBM Cloud Object Storage

Many organizations have already invested in enterprise-approved storage platforms and have stringent requirements around data residency, security, regulatory compliance, operational governance, and cost management. Requiring data to be stored in IBM Cloud Object Storage can create adoption challenges, additional costs, and data duplication across platforms.

I propose a Bring Your Own Storage (BYOS) capability that allows customers to use their own cloud storage accounts rather than mandatorily storing data in IBM Cloud Object Storage.

Supported storage platforms could include:

• Amazon S3
• Azure Blob Storage

Benefits:

• Eliminates unnecessary data duplication.
• Reduces storage and data transfer costs.
• Supports enterprise data residency requirements.
• Aligns with organizational security and compliance standards.
• Enables customers to leverage existing cloud storage investments.
• Simplifies data platform architecture and operations.
• Provides greater flexibility in multi-cloud environments.

How it should work:

• Allow administrators to register and manage external storage locations.
• Enable projects, catalogs, governance artifacts, and data products to reference customer-managed storage.
• Support storage-level encryption and customer-managed keys.
• Allow policy-based selection of storage locations based on region, domain, or data classification.
• Support seamless integration with existing security and access control models.
• Provide the option to choose between IBM-managed storage and customer-managed storage during environment setup.

Current Limitation / Use Case
Organizations adopting the platform may be required to use IBM Cloud Object Storage for certain workloads. However, many enterprises already maintain centralized cloud storage strategies using Amazon S3, Azure Data Lake Storage, or other approved storage platforms. Requiring data replication into IBM-managed storage introduces additional data movement, operational overhead, governance complexity, and storage costs.