I have been trying to find a diagram that shows how the SailPoint platform is deployed in the cloud, but I have only been able to find the VA (Virtual Appliance) diagram.
Does anyone know of an official diagram or documentation that covers the following?
Major components (compute, queues, streaming, databases, etc.) and how they connect
Data flow direction, including where customer data is stored and processed, and any segments that leave the environment
LLM/AI call path (which models and services are used, and whether a proxy layer exists)
As far as I know, SailPoint does not publicly provide a detailed cloud architecture diagram that exposes internal platform components such as compute services, databases, message queues, streaming platforms, AI service call paths, or tenant isolation implementation details. Those types of diagrams are typically not included in public product documentation for security reasons.
However, you may find useful information in the following resources:
SailPoint whitepapers and architecture-related documentation available through the Resource Library. [sailpoint.com]
Virtual Appliance (VA) architecture documentation, which explains the customer-managed connectivity component.
Regarding your specific questions:
Internal cloud components (compute, queues, databases, streaming, etc.): I have not seen an official public diagram that provides this level of implementation detail.
Data flow and tenant isolation: Some high-level information is available through SailPoint security, privacy, and trust documentation, but the underlying implementation details are generally not publicly documented.
LLM/AI architecture and model call paths: I have not seen public documentation describing the exact models, proxy layers, or internal AI service architecture.
PrivateLink and external integrations: Documentation exists for specific integration patterns and connectivity options, but not as a complete end-to-end platform architecture diagram.
Regarding SailPoint Human Fabric (SHF):
SHF is essentially the evolution/rebranding of Identity Security Cloud (ISC).
SailPoint’s developer documentation states that the platform was previously called IdentityNow, then Identity Security Cloud (ISC), and is now called SailPoint Human Fabric (SHF), while the product itself remains the same. [developer…lpoint.com]
SailPoint also describes Human Fabric as the evolution of Identity Security Cloud and part of the broader SailPoint Identity Security platform. [sailpoint.com], [investor.s…lpoint.com]
The ISC Microservices Architecture article confirms that ISC uses containerized microservices that can be deployed and scaled independently.
The SaaS Connectivity docs are useful for the connector and data-flow side, especially VA-based versus SaaS connectors.
For hosting, tenant security, encryption, and compliance, I would refer to the Trust Center and Trust Portal.
For the AI path, the AI at SailPoint page says generative AI primarily uses Anthropic Claude and Amazon Nova through Amazon Bedrock, with the models deployed in a privately operated AWS account that Amazon, not the model provider, controls, and that none of its AI features currently use customer data to train models across customers.
For the deeper pieces you mentioned, such as the exact AWS services, queues, databases, VPC and account boundaries, PrivateLink usage, tenant-isolation implementation, or whether there is an internal proxy in front of Bedrock, I would not try to infer those from the public docs. If this is for an architecture or security review, I would ask your SailPoint CSM, account team, or Field Security for that level of detail.
Also, SHF and ISC are not separate products. SailPoint’s developer docs explain that the platform was previously IdentityNow, then Identity Security Cloud, and is now SailPoint Human Fabric. It is the same platform with a newer name.