Salesforce, Dynamics CRM, and Boomi: Building a Scalable Integration Backbone

Enterprise systems fail when they don't communicate effectively. Sales, service, finance, operations, and analytics teams depend on real-time data flowing across platforms — yet enterprises rarely standardise on a single CRM or ERP. They operate hybrid ecosystems with Salesforce for sales automation, CPQ, and partner management, Dynamics 365 for finance, operations, and customer service, plus billing, marketing, and analytics systems.

Without proper integration, teams resort to point-to-point APIs, custom scripts, and manual reconciliation — approaches that create data silos, increase error rates, and don't scale. This guide covers how to architect a Boomi-based integration backbone connecting Salesforce and Dynamics CRM with enterprise-grade reliability, security, and observability.

Design a future-proof integration architecture:

• Boomi as iPaaS: Boomi provides the control plane governing how data moves, transforms, and triggers workflows. Its cloud-native runtime supports event-driven and batch integrations, hybrid/multi-cloud connectivity, low-code orchestration with enterprise controls, and built-in monitoring with retry logic — purpose-built for connecting SaaS ecosystems.
• Layered Architecture: Separate systems of record (Salesforce, Dynamics) from interaction systems (portals, mobile) and insight systems (analytics, BI). Boomi orchestrates data flows between layers — each system remains authoritative for its domain while consuming enriched data from others.
• API-Led Connectivity: Structure integrations in three API layers — Experience APIs (frontend-facing), Process APIs (business logic orchestration), and System APIs (direct system connectors). Boomi processes implement the Process layer, abstracting system-specific complexity from consumers.
• Event Mesh: Deploy Boomi Event Streams to create an event mesh — Salesforce publishes domain events (LeadConverted, OpportunityClosed), Dynamics publishes operational events (OrderFulfilled, InvoicePosted), and downstream consumers subscribe without direct coupling. This eliminates polling and enables near-real-time data flow.
• Connector Strategy: Use Boomi's native Salesforce and Dynamics 365 connectors for standard CRUD operations. Build custom connectors for specialised APIs (Salesforce Composite, Dynamics Batch). Version connectors independently to isolate platform API changes from integration logic.

Establish data governance across CRM and ERP boundaries:

• Golden Record Strategy: Define system-of-record rules per entity — Salesforce owns Leads, Opportunities, and Campaigns; Dynamics owns Invoices, Purchase Orders, and Inventory. Accounts/Contacts require a mastering strategy: one system is authoritative, and the other subscribes to changes with conflict resolution rules.
• Canonical Data Model: Define a platform-neutral data model in Boomi — map Salesforce objects and Dynamics entities to canonical schemas. This decouples systems from each other — when Salesforce adds fields or Dynamics changes entity structures, only the system-specific mapping changes, not the integration logic.
• Identity Resolution: Map Salesforce Record IDs to Dynamics GUIDs through a cross-reference table in Boomi. Handle merge scenarios (duplicate accounts merged in Salesforce) by cascading ID updates to Dynamics. Use external ID fields in Salesforce for upsert operations referencing Dynamics keys.
• Data Quality: Implement data validation in Boomi processes — reject records with missing required fields, invalid formats, or business rule violations before they reach target systems. Log rejected records for review and create Salesforce Tasks or Dynamics Activities for data stewards to resolve.
• Historical Data Migration: Plan initial data load separately from ongoing sync — Boomi batch processes handle bulk migration with progress tracking, deduplication, and reconciliation reports. Run migration in phases: reference data first (accounts, products), then transactional data (opportunities, orders).

Implement proven integration patterns for CRM-ERP connectivity:

• Customer 360: Synchronise account and contact data bidirectionally with mastering rules — Salesforce creates the account, Dynamics enriches with financial data (credit limit, payment terms, billing address). Boomi merges updates with last-writer-wins or field-level precedence rules to maintain a golden record across both systems.
• Lead-to-Cash Automation: When a Salesforce Opportunity closes, Boomi triggers creation of a Dynamics Sales Order, maps line items to Dynamics products, applies pricing from the ERP price list, and initiates the fulfilment workflow. Status updates flow back — shipping confirmation in Dynamics updates the Opportunity stage in Salesforce.
• Service-to-Billing Alignment: Salesforce Service Cloud cases trigger billing adjustments in Dynamics — credits, refunds, or subscription changes. Boomi validates the financial impact, applies approval workflows for amounts exceeding thresholds, and posts journal entries in Dynamics Finance automatically.
• Event-Driven Reactions: Publish domain events like "OrderCreated", "PaymentReceived", or "ContractRenewed" through Boomi Event Streams. Downstream consumers (analytics, notification services, third-party systems) subscribe independently — adding new consumers requires zero changes to publishers.
• Batch Synchronisation: Schedule nightly reconciliation jobs that compare Salesforce and Dynamics records, identify drift, and generate exception reports. Use Boomi's batch processing for high-volume operations — product catalogue sync, price list updates, and historical data backfills.

Build resilient integrations that handle failures gracefully:

• Retry Policies: Configure exponential backoff for transient failures (API rate limits, network timeouts) — start at 1 second, double each retry, cap at 5 minutes. Distinguish transient errors (retry) from permanent errors (dead-letter) based on HTTP status codes and error messages. Set maximum retry counts to prevent infinite loops.
• Dead Letter Queues: Route permanently failed messages to dead-letter queues with full context — original payload, error details, timestamp, and retry history. Create monitoring dashboards showing dead-letter queue depth, age of oldest message, and failure categories. Build UI for operations teams to inspect, fix, and replay failed messages.
• Idempotency: Design all integration operations to be idempotent — replaying a message should produce the same result as processing it once. Use Salesforce External IDs for upsert operations, Dynamics alternate keys for deduplication, and Boomi process properties for tracking processed message IDs.
• Circuit Breaker: Implement circuit breaker patterns — when error rates exceed thresholds (e.g., 50% failure over 5 minutes), stop sending requests to the failing system. Queue messages locally until the circuit resets and the target system recovers. Alert operations teams immediately when circuits open.
• Compensation Transactions: For multi-step integrations (create order → allocate inventory → charge payment), implement compensation logic that reverses completed steps when a later step fails. Boomi processes track step completion and execute compensating actions in reverse order.

Protect sensitive data flowing between enterprise systems:

• Authentication: Use OAuth 2.0 with client credentials for Salesforce Connected Apps and Azure AD for Dynamics 365 API access. Rotate credentials automatically — Boomi's credential management stores tokens securely and refreshes before expiration. Never pass credentials in integration payloads.
• Data Encryption: Encrypt data in transit (TLS 1.3) and at rest (Boomi encrypted process properties). For sensitive fields (SSN, credit card numbers, health records), apply field-level encryption in Boomi before writing to target systems. Use Salesforce Shield Platform Encryption and Dynamics Customer-Managed Keys for storage encryption.
• Access Control: Implement role-based access to Boomi processes — integration administrators manage configurations, developers build processes, and operators monitor execution. Use Salesforce Named Credentials and Dynamics Application Users with minimal required permissions for integration service accounts.
• Audit Trails: Log all data movement — what data was read, transformed, and written, by which process, at what time. Store audit logs for compliance retention periods (7 years for SOX, 6 years for HIPAA). Boomi's process reporting provides execution history; supplement with custom audit records in both CRM systems.
• Data Residency: Ensure integration processing respects data sovereignty requirements — configure Boomi Atom locations to process data within required geographic boundaries. Validate that Salesforce Shield and Dynamics data residency settings align with integration routing to maintain regulatory compliance.

Maintain operational visibility across the integration landscape:

• Boomi Dashboard: Monitor process execution metrics — success rate, average duration, throughput, and error frequency. Set SLA-based alerts — if Lead-to-Cash processing exceeds 30 seconds or error rate exceeds 2%, trigger notifications to integration team channels (Slack, PagerDuty).
• Health Checks: Implement synthetic transactions that verify end-to-end connectivity every 5 minutes — create test records in Salesforce, verify they appear in Dynamics within SLA, and clean up. Alert on missing or delayed propagation before business users notice data gaps.
• Capacity Planning: Track API consumption against Salesforce daily API limits and Dynamics throttling thresholds. Boomi batch operations should schedule during off-peak hours to preserve API capacity for interactive users. Project capacity needs based on data volume growth trends.
• Log Aggregation: Centralise integration logs in an observability platform (Datadog, Splunk, ELK) — correlate Boomi process logs with Salesforce debug logs and Dynamics trace logs for end-to-end troubleshooting. Use correlation IDs that flow through all systems for distributed tracing.
• Business Metrics: Beyond technical metrics, track business outcomes — order-to-cash cycle time, data freshness (how quickly Salesforce changes appear in Dynamics), and exception rates per integration pattern. These metrics validate that the integration backbone delivers business value.

A well-architected integration backbone transforms disconnected CRM and ERP systems into a unified enterprise platform. Key principles:

• Treat integration as a product — invest in architecture, monitoring, and documentation as seriously as application development.
• Define data ownership — establish system-of-record rules before building any integration processes.
• Design for failure — implement retry, dead-letter, circuit breaker, and compensation patterns from day one.
• Monitor business outcomes — technical metrics alone don't validate integration success; track cycle times and data freshness.

MetaDesign Solutions provides expert enterprise integration services — from Salesforce-Dynamics CRM architecture design and Boomi implementation through data modeling and master data governance, event-driven integration patterns, security and compliance configuration, production monitoring setup, and ongoing managed integration services for hybrid CRM/ERP ecosystems.