Skip to main content

Systems operational

Existing client enquiries
NetEvolution
Insight / System Integration

Salesforce to SITS integration: event-driven patterns with Azure Service Bus

When a university connects Salesforce CRM to Tribal SITS, synchronous point-to-point APIs inevitably buckle under peak load. Here is how an event-driven architecture guarantees zero message loss and real-time visibility.

01 / the structural friction

Two philosophies, one student lifecycle

Modern UK universities increasingly run Salesforce (typically Education Cloud or the Education Data Architecture / EDA) to power applicant recruitment, marketing automation, and student support desks. Concurrently, Tribal SITS:Vision remains the statutory system of record for registry, enrolments, fees, and degrees.

The conflict arises because these systems operate on fundamentally different paradigms. Salesforce is a modern, cloud-native CRM expecting instantaneous webhook triggers and flexible JSON objects. SITS is a rigorous on-premise or hosted relational engine built around transactional batch integrity, strict academic sessions, and scheduled nightly reconciliations.

When institutions attempt direct, synchronous HTTP integrations between them, failure is virtually guaranteed during high-volume events such as Confirmation & Clearing or term registration. If SITS slows down under heavy database load, Salesforce HTTP timeouts trigger cascading retries, exhausting API quotas, locking database tables, and generating orphaned records.

02 / core architecture

The decoupled event-driven pattern

The proven solution deployed across enterprise institutions is an asynchronous message broker pattern built on Azure Service Bus. Rather than allowing systems to call each other directly, every state change emits a durable message into an isolated topic.

When an applicant accepts an offer in Salesforce, a Change Data Capture (CDC) event or outbound message is published to the bus. Downstream microservices consume the event, perform schema transformation, validate academic prerequisites against SITS rules, and queue the record for insertion.

Because the queue acts as an elastic buffer, traffic spikes that would otherwise crash SITS interfaces are safely smoothed. If SITS undergoes planned maintenance or suffers transient connectivity issues, the messages wait safely in Azure with transactional persistence — ensuring zero lost applications.

03 / resilience requirements

Three non-negotiable patterns for HE integration

  • Idempotent processing: Every message carries a unique transaction hash (e.g. UCAS ID + course code + timestamp). If a network glitch causes redelivery, the integration worker recognises the duplicate and ignores it.
  • Dead-letter queue (DLQ) triage: Poison messages or malformed records are automatically routed to a DLQ rather than blocking the entire queue. Dedicated worker alerts notify integration engineers with exact payload diagnostics.
  • Near-real-time telemetry dashboard: An operations dashboard displaying queue depths, processing latencies, dead-letter counts, and active worker statuses — giving registrars and IT directors immediate visibility.

04 / verified implementation

Proven on the University of London Global Digital Campus

This architecture is not theoretical. NetEvolution's founder is currently the integration developer delivering this exact pattern for the University of London's Global Digital Campus programme — connecting Salesforce, SITS student records, and the Canvas learning management platform with a custom observability portal for real-time queue recovery.

By isolating the transactional interfaces behind Azure Service Bus, the institution has eliminated synchronisation race conditions and enabled automated failover, supporting thousands of global distance learners across multiple federated colleges.

05 / strategic takeaways

Moving towards governed agentic integration

Once an event-driven integration backbone is established, it becomes the launchpad for agentic workflow automation. Intelligent agent workers can subscribe to message queues to triage complex exception cases, draft contextual communications for student advisers, and reconcile discrepancies without human intervention.

To explore how your institution can resolve CRM and student record disconnects, explore our infrastructure integration services or read our guide to SITS:Vision automation.

Sources

Eliminate the integration gap between Salesforce and SITS

We audit existing point-to-point connections, identify failure points, and deliver a production-hardened Azure Service Bus blueprint tailored to your university's data schemas.

Request an architecture review