In a complex landscape where multiple systems contain overlapping customer data, each with its own primary key, the core architectural challenge is Identity Management. To ensure that an update in Salesforce (the System of Record) correctly updates "Customer A" in the ERP and "Customer A" in the Data Lake, a Master Data Management (MDM) strategy is required.
An MDM solution creates a Cross-Reference (X-Ref) Table or a "Golden Record" that maps the unique identifiers from all systems. In the Salesforce record, the architect should implement External ID fields for each corresponding system (e.g., ERP_ID__c, Ticket_System_ID__c).
Why this is the superior recommendation:
Bidirectional Integrity: When the middleware receives an update from the ERP, it uses the ERP_ID__c to perform an "upsert" in Salesforce, ensuring no duplicates are created.
Traceability: It allows for easy auditing of data lineage across the enterprise.
Decoupling: Salesforce doesn't need to know the internal logic of the ERP; it simply holds the reference key.
Option B (CDC) is a delivery mechanism, not an identity management strategy; it tells you that something changed, but not which record in the ERP it corresponds to without the ID mapping. Option C (Local caching in middleware) is an "anti-pattern" because it makes the middleware stateful; if the middleware cache is lost or out of sync, the entire integration breaks. By designing an MDM-based mapping solution directly within the data model, the architect ensures a robust, scalable, and transparent identity framework for the entire enterprise.

When an organization must interact with multiple external services (34 different shipping providers) that change frequently, the primary architectural goal is to decouple Salesforce from the specific endpoints and protocols of those services.11 The architect should propose us12ing middleware to abstract the call. By introducing a middleware layer (such as an ESB or iPaaS), Salesforce makes a single, standardized call to the middleware, passing the country and order details. The middleware then handles the "Process Orchestration":
Service Routing: Determining which shipping service is valid for the specified country.
Protocol Mediation: Translating the standard Salesforce request into the specific format (REST, SOAP, XML, etc.) required by each individual shipping provider.
Maintenance: When a shipper is added or removed, the change is made in the middleware without requiring any code or metadata updates in Salesforce.
Option A (Picklists) creates a high administrative burden and does not solve the underlying technical complexity of making callouts to dozens of different APIs. Option B (Platform Events) is asynchronous and would introduce unnecessary latency for a sales rep who needs an estimate "on the spot" to complete an order. Middleware abstraction provides a scalable, maintainable, and agile architecture that allows NTO to optimize its shipping operations without disrupting the CRM user experience.

When designing a solution that requires Salesforce to receive and display updates from a back-end ERP (such as order status changes), message durability is a critical factor for ensuring data consistency. In an event-driven architecture using Platform Events or Change Data Capture (CDC), Salesforce utilizes an event bus to handle these incoming notifications.
For high-volume event messages, the Salesforce platform provides a native 72-hour (3-day) retention window. This is a significant architectural advantage for several reasons:
System Resilience: If the Salesforce org or the integration middleware experiences a temporary disruption or is undergoing maintenance, the event messages published by the ERP remain stored in the bus for up to 3 days.
Data Recovery: Once the connection is restored, the subscribing system (Salesforce) can use the Replay ID to catch up on any missed events from the last 72 hours, ensuring that order stages like "Order Shipped" or "Order Paid" are not missed.
SLA Management: This 3-day window exceeds the 24-hour limit of older technologies like PushTopics or Outbound Messaging (Option A), providing more breathing room for disaster recovery scenarios.
While ReplayID -1 (Option C) is used to subscribe only to new events published after the subscription starts, it does not address the durability or retention of historical events needed for recovery. By highlighting the 72-hour retention window, the integration architect provides a design that is robust against outages and guarantees that the "System of Engagement" (Salesforce) stays synchronized with the "System of Record" (ERP).

For HR synchronization requiring high reliability and a multi-day data retention window, Change Data Capture (CDC) is the optimal architectural choice.12 CDC is an event-driven integration framework that automatically publ3ishes change events for Sale4sforce records (Create, Update, Delete, and Undelete). A critical feature of CDC for failure recovery is its 72-hour (3-day) retention window. If the external HR system or the replication middleware experience downtime, the change events remain stored in the Salesforce event bus for up to three days. Once the HR system is back online, the subscriber can use the Replay ID of the last successfully processed event to retrieve all missed notifications from the bus, ensuring zero data loss during the outage.
In contrast, Outbound Messaging (Option B) only provides a 24-hour retry window. If the target system is unavailable for more than one day, the notifications are dropped, making it unsuitable for a 3-day requirement. Callouts (Option C) are typically synchronous and do not have native, platform-managed storage or retry mechanisms for failed requests; once a callout fails, the data is lost unless a custom retry architecture is built. CDC's native, durable event bus provides the necessary scalability and resilience for critical HR data synchronization without requiring complex custom code for error handling and data recovery.

In a synchronous Request-Reply integration-where a bank agent is waiting for a real-time credit check to open an account-the error handling strategy must balance user experience with system resilience. Handling these errors at the Middleware layer is the architecturally preferred solution for managing complex retry logic and providing a clean response to Salesforce.
If the external credit agency's service is momentarily unavailable, the middleware (such as an ESB or MuleSoft) can automatically retry the request multiple times using a pre-defined strategy (e.g., exponential backoff). This "self-healing" behavior can often resolve transient network issues before the Salesforce agent even realizes there was a problem. If the retries fail, the middleware then returns a structured error message to Salesforce, which is displayed to the agent via the UI.
Option B (Fire and Forget) is unsuitable for this use case because the agent needs the result immediately to proceed with the bank account opening; they cannot afford to wait for a background process to finish hours later. Option C (Mock Service) is a testing tool and has no place in a production environment where real financial decisions are being made. By delegating error management to the middleware, UC ensures that its Salesforce instance remains performant (avoiding long-running request timeo1112uts) while maximizing the chances of a successful credit check through automated, controlled retries.1314