A curated subset of the article library focused on architecture, system design, resilience, scalability, event-driven systems and production engineering trade-offs.
Search the collection or filter by the topics represented in these longer-form architecture and engineering articles.
A practical architecture pattern for isolating failed integration work, preserving appropriate recovery context, and enabling safe replay without turning transient failures into permanent data loss.
Read deep dive →A practical architecture guide to API governance across design standards, ownership, contracts, security policies, traffic controls, observability and lifecycle management.
Read deep dive →Experience, Process and System APIs are architectural responsibilities, not a requirement to add three network hops to every integration.
Read deep dive →A production architecture guide to protecting event consumers and downstream systems when event arrival rates exceed processing capacity.
Read deep dive →Event producers and consumers evolve independently. This deep dive explores additive change, semantic compatibility, versioning and migration without turning the event bus into a coordination bottleneck.
Read deep dive →A production design for consuming high-volume Salesforce Change Data Capture events without depending on the 72-hour event-bus window: durable handoff, replay checkpoints, sequencing, parallel processing, backpressure and recoverable error handling.
Read deep dive →A practical architecture guide to preventing duplicate business effects when APIs, events, batches, and recovery workflows are retried or replayed.
Read deep dive →A practical architecture for tracing business transactions across APIs, events, queues and downstream systems without confusing logs with observability.
Read deep dive →A practical architecture guide to reliable event publication and consumption when business state, brokers, databases, and external systems cannot share one atomic transaction.
Read deep dive →A practical architecture for consuming Salesforce CDC and Platform Events reliably, with replay checkpoints, idempotency, recovery paths, observability, and downstream failure handling.
Read deep dive →Retries are only one recovery mechanism. This deep dive separates retry, replay and reprocessing and shows how to design each without creating duplicate business effects.
Read deep dive →A practical guide to choosing MuleSoft processing patterns for large datasets, including batch jobs, aggregation, concurrency, memory, API limits, failure recovery, and observability.
Read deep dive →Articles also includes tutorials, Quick Reads and implementation notes across Salesforce, MuleSoft, DataWeave and integration engineering.