.png)
Faire Integration Guide
Connect Faire’s approved partner REST APIs and selected webhook events with ERP, commerce, fulfillment, and business systems through Martini workflows.
Faire integration options at a glance
Faire’s primary integration model is an authenticated partner REST API for approved applications and integrations. Depending on account role and permissions, API access can support Products, Product variants, inventory, Orders, Shipments, Brands, and Retailers. Faire also supports webhook-style notifications for selected partner events, although event coverage and delivery behavior must be confirmed for each integration. Martini can consume Faire REST endpoints, expose an HTTP endpoint for supported notifications, schedule incremental synchronization, and map Faire data into ERP, commerce, fulfillment, accounting, or internal systems. Credentials are stored as environment-managed secrets, while pagination, rate limits, retries, idempotency, and reconciliation are handled within workflows.
Common Faire integration patterns
Common Faire data objects used in integrations
Authentication and security considerations
Partner approval and permissions
Faire API access may require approval or onboarding. Confirm the account role, enabled API products, object permissions, read and write access, and production credential requirements before implementation.
Credential management
Faire requests require authenticated partner credentials. Depending on the application program, access tokens, API keys, or OAuth-based authorization may apply. Store these values as Martini environment-managed secrets rather than embedding them in workflows.
Webhook protection
For selected Faire notifications, validate the authentication or signature mechanism documented for the partner integration before processing. Use HTTPS, event identifiers, replay protection where supported, and record the notification safely before acknowledging it.
Operational considerations for Faire integrations
Pagination and incremental retrieval
Process Product and Order collections using Faire’s documented pagination. Persist cursors, timestamps, or equivalent checkpoints where available, and account for records changing during a long-running synchronization.
Rate limits and retries
Handle HTTP 429 responses, transient 5xx errors, timeouts, and retry-after guidance when provided. Use bounded exponential backoff and route exhausted failures to reconciliation rather than repeatedly replaying writes.
Idempotency and reconciliation
Scheduled polling and webhook delivery can overlap. Use stable Faire identifiers, event IDs, or object/version keys to prevent duplicates. Reconcile Orders, Shipments, Products, and inventory against target systems to recover from missed notifications.
Schema and version changes
Isolate Faire-specific mappings and monitor API versions, required fields, enum values, Product variant structures, and event payload changes. Validate required fields before sending data downstream and preserve unknown fields where practical.
Testing and observability
Test with the permissions and object coverage of the actual partner account. Monitor workflow logs, API response patterns, processing checkpoints, and reconciliation outcomes.
Why use Martini instead of scripts or point-to-point integrations?
Orchestration instead of isolated scripts
Martini provides a maintainable workflow layer for calling Faire APIs, receiving selected events, transforming data, applying business rules, and coordinating multiple target systems. This avoids duplicating authentication, pagination, retries, and mapping logic across scripts.
Reusable integration assets
Faire-specific API handling, validation, identifier cross-references, and error paths can be reused across order, catalog, inventory, and shipment workflows. Martini can also expose controlled APIs for downstream applications.
Operational control
Workflows can combine scheduled synchronization with event-driven processing, persist checkpoints, handle failures, and support reconciliation. This provides a clearer operational model than unmanaged point-to-point jobs while retaining flexibility for custom logic.