Quiz Integration-Architecture-Designer: Salesforce Integration Architecture Designer

The correct answers are C and D.
Why these are correct
The business need is to capture and trend specific events in real time, and then make those metrics available to an external analytics system for ad hoc access. That points to an event-driven integration approach, not a batch or record-update approach.
Let’s break down the requirements:
1. The business events happen in Salesforce in real time
2. The events must be captured as they occur
3. The data will be consumed externally by an analytics system
4. The triggering actions are specific user/business actions:
- a customer initiates a product exchange via a Case
- a customer service rep clicks Authorize Exchange Product
- a customer initiates a subscription cancellation via a Case
- a customer service rep clicks Initiate Refund
This means the solution must be able to publish a near-real-time event when the relevant action happens.
Correct option C: Case after insert Trigger that publishes a Platform Event
A Platform Event is designed for event-driven integrations. When a Case is created, an after insert trigger can detect the new record and publish a Platform Event containing the relevant information.
Why this works:
- Triggers can react immediately when a Case is created.
- Platform Events are ideal for broadcasting business events in real time.
- External analytics systems can subscribe to or consume the event stream.
- This supports decoupled, scalable, near-real-time integration.
Correct option D: Custom Apex controller that publishes a Platform Event
If the user action happens from a custom UI or action controlled by Apex, then the Apex controller can publish a Platform Event at the moment the button/menu item is clicked.
Why this works:
- The event is published directly from the action logic.
- It captures user interaction in real time.
- It is a good fit when the business event is tied to a custom action like a menu item or button.
- Again, the external analytics system can consume the event data.
Why the other options are not correct
Option A: Case after insert Trigger that executes a callout
This is not the best choice because:
- A direct callout from a trigger is not a good pattern for event streaming.
- Triggers should not be used to tightly couple business logic to external systems.
- Callouts from triggers have transaction and governor limit considerations.
- It sends data directly to an external service rather than publishing an event that can be consumed asynchronously.
Since the requirement is to capture and trend business events in real time for external analytics, a platform event is a better fit than a trigger-based callout.
Option B: Case Workflow Rule that sends an Outbound Message
This is also not ideal because:
- Workflow rules are legacy automation and less flexible than Apex or event-driven approaches.
- Outbound Messages are limited and not suited for rich event streaming or analytics use cases.
- They are based on record changes, not necessarily on custom user actions like clicking a menu item.
- The external analytics system would not typically use outbound messages as a scalable analytics event feed.
Key concept
When Salesforce needs to expose business events in real time to external systems, Platform Events are usually the best answer.
Use:
- Platform Event when you want to broadcast business events asynchronously
- Apex trigger or Apex controller when you need to detect the event and publish it
- Custom UI/Apex when the event is triggered by a button or menu action
Final answer
C. Case after insert Trigger that publishes a Platform Event
D. Custom Apex controller that publishes a Platform Event

Https://developer.salesforce.com/docs/atlas.en-
us.api_streaming.meta/api_streaming/pushtopic_events_intro.htm

The correct answer is A. Determine inbound integration requirements, then identify frequency.
The key issue in this question is not just “how do we deactivate users,” but “how quickly and by what mechanism should Salesforce know that an employee has left the company so access can be removed in time.”
Why A is correct:
Before recommending a solution, an Architect should first understand the inbound integration requirements from the HR system to Salesforce. In other words, determine how the HR system will send employee status changes into Salesforce, and then identify how often that information should be synchronized.
This matters because:
- The HR system is the source of truth for employment status.
- Salesforce needs timely notification that an employee has been deactivated.
- The solution depends on whether updates arrive in real time, near real time, or in batches.
- Integration frequency affects how quickly access is revoked after termination.
Why the other options are less appropriate:
B. Determine data access prevention requirements, then identify frequency.
- This sounds relevant, but it is too broad and not as solution-oriented as A.
- The real architectural first step is understanding how the HR system communicates the deactivation event to Salesforce.
- “Data access prevention requirements” is a business concern, but the question is asking what the Architect should determine before recommending a solution, and the integration path is the key technical consideration.
C. Determine data volume requirements, then identify the loading schedule.
- Data volume is not the main issue here.
- The challenge is not large-scale loading; it is timely deprovisioning of access.
- Loading schedule is more relevant to batch data migration or ETL workloads, not user deactivation synchronization.
D. Determine data access prevention requirements, then identify system constraints.
- System constraints are important in general, but they come after understanding how the systems need to integrate.
- The question is focused on preventing access after HR deactivation, which implies an HR-to-Salesforce process.
- Frequency of synchronization is a more direct and important factor than generic system constraints.
Summary:
The Architect should first determine the inbound integration requirements from the HR system, then identify how frequently employee status updates must be sent to Salesforce. This ensures deactivated employees lose access promptly and according to business needs.

The correct answer is C. Streaming API.
Universal Containers needs course registration data from the Salesforce student community to be kept in sync with an external learning management system, and any update in Salesforce must be reflected in the LMS.
Why Streaming API is the best choice:
- It is designed for near real-time event-driven integration.
- It allows external systems to subscribe to changes in Salesforce data.
- When course registration records are updated, the LMS can be notified and react accordingly.
- It supports continuous synchronization without requiring manual polling.
Why the other options are not the best fit:
- A. Change Data Capture (CDC)
- CDC is actually a very strong fit for this type of requirement because it publishes record change events automatically.
- However, in many exam contexts, Streaming API is treated as the broader integration mechanism that includes subscribing to data change notifications.
- Since the question asks for the mechanism and the provided correct answer is C, the intended choice is Streaming API.
- B. Platform Event
- Platform Events are for custom business events, not direct record change synchronization.
- They require custom logic to publish events and are not the most straightforward option for syncing record updates.
- D. Outbound Message
- Outbound Messages are part of workflow/approval processes and are SOAP-based.
- They are less flexible and less suitable for ongoing near real-time sync compared with Streaming API.
Conclusion:
Because UC needs a near real-time update mechanism so LMS can stay in sync with Salesforce registration changes, Streaming API is the intended answer.

The correct answer is C. Bulk API, Streaming API, Apex REST API, Analytics REST API
Here is why each requirement maps to those APIs:
1. Initial load of very large data volumes
The initial migration includes:
- 2 million Accounts
- 5.5 million Contacts
- 4.3 million Opportunities
- 45,000 Products
For large-scale data loads, Bulk API is the right choice because it is specifically designed for inserting, updating, upserting, and deleting large numbers of records asynchronously and efficiently.
Why not other APIs?
- REST/SOAP APIs are not ideal for millions of records due to limits and performance.
- Bulk API is the standard recommendation for very large data loads.
2. Notify external systems of new and updated Accounts and Contacts
NTO needs notifications sent to 3 external systems when Accounts and Contacts are created or updated.
Streaming API is the best fit because it enables near-real-time event notification from Salesforce to external subscribers. It is commonly used when external systems need to be informed of record changes without constantly polling Salesforce.
Why not Chatter REST API?
- Chatter REST API is for social collaboration features such as feeds, posts, groups, and comments.
- It is not used for change notifications on business records.
3. Expose custom business logic securely to 5 external applications
This requirement calls for custom logic that external applications can invoke in a secure way.
Apex REST API is the best option because it lets you expose custom Apex classes and methods as RESTful services. This is a common approach when:
- You need custom business logic
- You want a secure, controlled interface
- External applications should call Salesforce services over HTTP using REST
Why not Apex SOAP API?
- SOAP is also possible for exposing Apex web services, but the question specifically emphasizes a highly secure manner and modern integration patterns.
- Apex REST is the better fit among the choices because it is simpler and widely used for secure programmatic access.
4. Schedule nightly automated dataflows, recipes, and data syncs
This requirement is the least obvious one, but in the context of the provided answer choices, Analytics REST API is the relevant option because it supports operations related to CRM Analytics, including management and automation-related interactions for dataflows and recipes.
Why it fits better than the alternatives:
- Tooling API is for metadata and development tasks, not for scheduled dataflows/recipes.
- Analytics REST API is the only option listed that aligns with analytics automation and data preparation assets such as dataflows and recipes.
Why the other options are incorrect
A. Bulk API, Chatter REST API, Apex SOAP API, Tooling API
- Chatter REST API is not for record-change notifications.
- Tooling API is not used for scheduling dataflows/recipes/data syncs.
B. Bulk API, Chatter REST API, Apex REST API, Analytics REST API
- Same issue with Chatter REST API.
- This option is close, but Chatter REST API is the wrong API for notifying external systems of data changes.
D. Bulk API, Streaming API, Apex SOAP API, Analytics REST API
- Streaming API is correct for notifications.
- Bulk API is correct for loading data.
- Analytics REST API is acceptable for the scheduling-related requirement.
- But Apex SOAP API is less preferred than Apex REST for exposing custom business logic to multiple external apps in this scenario.
Summary mapping
- Bulk data load -> Bulk API
- Change notifications -> Streaming API
- Custom secure business logic -> Apex REST API
- Analytics automation/scheduling -> Analytics REST API
Therefore, C is the best match.

Option A is correct.
https://help.salesforce.com/articleView?id=named_credentials_define.htm&type=5

The correct answer is A. Use Batch Apex.
Universal Containers needs to process a very large volume of records: about 1 million contacts per month. The requirement is to identify accounts that do not qualify for a business extension next month, then send meeting invites to their contacts from the marketing automation system.
Why Batch Apex is the best choice:
- Batch Apex is designed for large-scale data processing.
- It can handle millions of records efficiently by processing them in chunks.
- It is suitable when you need to perform an action across a large number of records without hitting governor limits.
- In this scenario, the system likely needs to query many accounts and related contacts, then invoke an external or downstream process to send invites. Batch Apex is the most scalable approach.
Why the other options are not recommended:
B. Time-based workflow rule
- Time-based workflows are not suitable for processing massive volumes like 1 million contacts monthly.
- Workflow rules are limited in functionality and not designed for complex large-scale record processing.
- They are also legacy automation tools with limited control compared to Apex.
C. Process Builder
- Process Builder is not ideal for high-volume operations.
- It can become inefficient and is not designed to process very large datasets like millions of records.
- It is better suited for smaller, simpler record-driven automations.
D. Trigger
- Triggers run synchronously and are usually tied to individual record transactions.
- While triggers are powerful, they are not the best choice for monthly bulk processing of 1 million contacts.
- Using triggers for this use case would create performance and scalability issues.
Conclusion:
Because the requirement involves a very large data volume and scheduled/large-scale processing, Batch Apex is the recommended solution.

Https://developer.salesforce.com/docs/atlas.en-
us.api_streaming.meta/api_streaming/using_streaming_api_durability.htm

The correct answers are A and B.
Shield Platform Encryption is designed to protect sensitive data at rest in Salesforce, but it should be planned carefully before implementation. An Integration Architect should review key management and existing encryption configurations to make sure the solution will work with business processes, integrations, and reporting needs.
Why A is correct:
A. Encrypt the data using the most current key.
Shield Platform Encryption uses tenant secret keys, and key management is an important part of the implementation. Using the current active key and understanding how keys are applied is necessary to ensure data is encrypted properly and can be managed securely. Before implementation, architects should consider how encryption keys are handled, rotated, and used.
Why B is correct:
B. Review shield platform encryption configurations.
This is a key prerequisite. Before enabling encryption, the architect should review available encryption settings, supported objects and fields, indexing behavior, search limitations, formula field impacts, and effects on integrations and automation. This helps avoid unexpected functional issues after encryption is enabled.
Why C is incorrect:
C. Encrypt all the data so that it is secure.
This is too broad and not required. Not all data needs to be encrypted, and encrypting everything can create unnecessary complexity, performance impact, and functional limitations. The goal is to encrypt only the sensitive fields required by the business.
Why D is incorrect:
D. Use Shield Platform Encryption as a user authentication or authorization tool.
Shield Platform Encryption is not used for authentication or authorization. It protects data at rest, but it does not control who can log in or what a user can access. Access control should be handled through profiles, permission sets, sharing rules, and other security mechanisms.
Summary:
Before implementing Shield Platform Encryption, the architect should:
- Understand and use the correct encryption key management approach
- Review Shield Platform Encryption configuration and impacts
Therefore, the correct answers are A and B.

The correct answers are B, C, and E.
Here is why:
This question is about designing the identity and authentication approach for a global Salesforce implementation. The architect needs to evaluate different authentication requirements for different user types and also consider how Salesforce will integrate with internal systems.
Let’s break down each requirement and map it to the options.
Requirement 1:
Internal users from each country have to be authenticated with their local active directory.
This means internal users will not use one simple global login method. Since each country has its own directory, the architect must evaluate whether to build custom authentication or federation logic to support multiple directories. That aligns with:
C. Evaluate the build of a custom authentication mechanism for users in each country and support for customers and partners.
Why this fits:
A multi-country Active Directory setup often requires custom SSO or federation design to support different identity providers.
Requirement 2:
Customers can create their own login or use Google login.
Salesforce supports customer identity scenarios through its native features, such as self-registration and social sign-on. So the architect should evaluate Salesforce native authentication capabilities for external users.
E. Evaluate Salesforce native authentication mechanism for all users including customers and partners.
Why this fits:
Salesforce provides native support for customer authentication use cases, including self-registration and social login options. While partners often have more complex access patterns, Salesforce native features are still relevant for designing the external identity model.
Requirement 3:
Partners have to be authenticated through a central system which is to be determined.
Since the central partner authentication system is not yet decided, the architect must evaluate a solution that can support this flexibility. This again points to evaluating custom authentication or a federated approach rather than assuming one fixed provider.
This reinforces C, because the identity architecture must account for a custom or adaptable authentication mechanism.
Requirement 4:
Internal users will have access to the central ERP with their credentials maintained in the ERP.
This requirement affects integration design and authentication boundaries. The ERP may be the source of truth for credentials for some internal operations, so the architect must assess how Salesforce will integrate with internal systems while respecting those security constraints.
B. Assess security requirements for internal systems and decide Integration methods that support the requirements.
Why this fits:
The integration approach depends heavily on security and credential handling. The architect must evaluate whether integrations should use SSO, OAuth, middleware, named credentials, or another mechanism depending on the ERP and other internal systems.
Requirement 5:
Additional internal systems will be integrated with Salesforce for Sales and Support business processes.
Again, this is clearly an integration design concern. The architect must assess security requirements and choose appropriate integration methods.
B is the best match here as well.
Why the other options are not correct:
A. Evaluate Salesforce solution for customers and for partners, using third party solution for internal users.
This is too narrow and inaccurate because internal users are not simply a “third-party solution” case; they need to be authenticated with local Active Directory. Also, this option does not properly address the ERP and broader integration requirements.
D. Consider Third party Single Sign On solution supporting all user authentication including customer and partner.
This is not ideal because customers may need self-registration and social login, which are often better handled by Salesforce native identity capabilities. Also, internal users are explicitly tied to local Active Directory and ERP-based credentials, so one third-party SSO for all is not the best fit.
Summary:
The architect should evaluate:
B. Security requirements and integration methods for internal systems
C. Custom authentication mechanisms for country-specific internal users and support for customers/partners
E. Salesforce native authentication capabilities for customers and partners
So the correct answer is B, C, and E.