.png)
Sigma Computing Integration Guide
Integrate Sigma Computing with enterprise systems through REST APIs, scheduled workflows, selective automation features, and analytical exports.
Sigma Computing integration options at a glance
Sigma Computing’s primary programmatic integration mechanism is its REST API, which supports platform and administrative operations involving Workbooks, Datasets, Connections, Users, Teams, and selected exports. Sigma uses API clients and bearer access tokens for authentication. Selected action and automation features may provide callback-style behavior, but broad webhook coverage for all object changes is not confirmed, so scheduled polling is the safer synchronization pattern. Export operations can be asynchronous and may produce CSV, Excel, PDF, or image files where supported. Martini can orchestrate these calls, map JSON, poll long-running operations, deliver files, and expose a controlled API façade.
Common Sigma Computing integration patterns
Common Sigma Computing data objects used in integrations
Authentication and security considerations
API authentication
Sigma documents API clients and bearer access tokens for REST API authentication. The required permissions depend on the resource and operation, including tenant-level operations involving Users, Teams, Workbooks, or Connections.
Credential protection
- Use a dedicated Sigma API client rather than personal credentials.
- Store client credentials and tokens in Martini secrets or environment configuration.
- Apply least-privilege permissions and rotate credentials according to enterprise policy.
- Do not place access tokens, exported report contents, or sensitive payloads in workflow logs.
Tenant and data security
Confirm the customer’s regional API hostname and tenant configuration before deployment. Preserve Sigma permission boundaries when exposing operations through a Martini API and apply retention controls to exported files and temporary downloads.
Operational considerations for Sigma Computing integrations
Reliability and scale
- Confirm Sigma rate limits and implement exponential backoff for throttling and transient failures.
- Treat collection responses as paginated unless the endpoint explicitly states otherwise.
- Persist operation identifiers and poll asynchronous exports with a controlled interval and timeout.
- Use Sigma object identifiers and durable mappings to make retries idempotent.
Change management
Avoid undocumented response fields, validate required fields, tolerate additive fields, and monitor Sigma API release and deprecation information. Test representative tenant configurations, export formats, permissions, and download URL behavior before production deployment.
Observability
Log correlation identifiers, endpoint names, status codes, workflow execution identifiers, checkpoints, and operation states without logging credentials or sensitive report contents. Reconcile missed or unsupported events through scheduled polling.
Why use Martini instead of scripts or point-to-point integrations?
Beyond point-to-point scripts
Martini provides a maintainable workflow layer around Sigma REST APIs, scheduled polling, asynchronous exports, and downstream delivery. This separates authentication, vendor-specific mappings, business rules, retries, and operational state from individual scripts.
- Expose a controlled Martini API instead of distributing Sigma credentials across applications.
- Reuse mappings and orchestration for Users, Teams, metadata, and exports.
- Apply consistent validation, idempotency, retries, timeout handling, and audit logging.
- Coordinate Sigma with warehouses, governance systems, file destinations, and enterprise applications.
Appropriate system boundary
Use Sigma REST APIs for Sigma-specific administration and metadata. Use the underlying warehouse or database for high-volume analytical data access, avoiding unnecessary repeated exports through Sigma.