integration-automation-patterns 0.2.0

Reference patterns for reliable enterprise integration, workflow automation, and system-of-record synchronization.

integration-automation-patterns

Practical patterns for enterprise integration, workflow orchestration, and system-of-record synchronization in complex operating environments.

Why this repo exists

Enterprise modernization usually breaks down at the integration layer:

• brittle handoffs between systems
• inconsistent event handling
• weak retry and idempotency models
• workflow logic scattered across tools
• poor visibility into operational state

This repository is a public-safe reference for patterns that help teams build more reliable integration and automation systems. The patterns are platform-agnostic and cloud-agnostic — applicable across any combination of CRM, ERP, ITSM, and custom services, on any cloud environment (AWS, GCP, Azure, OCI) or on-premises.

Scope

This repo focuses on:

• event-driven integration patterns with explicit retry and idempotency models
• system-of-record synchronization with authority boundaries
• workflow orchestration and escalation boundaries
• observability for automation flows
• public-safe architecture notes for enterprise operations

The patterns do not assume any specific vendor, broker, or cloud platform.

Modules

Core reliability:

• event_envelope.py — reliable event transport with explicit delivery status, bounded retry, and audit logging. Works with any broker (Kafka, SQS, Azure Service Bus, GCP Pub/Sub, RabbitMQ).
• sync_boundary.py — bi-directional system-of-record sync with field-level authority assignment, conflict detection, and exclusion management.

Resilience patterns (v0.2.0):

• circuit_breaker.py — thread-safe CLOSED/OPEN/HALF_OPEN state machine with automatic recovery probing and .call decorator.
• saga.py — distributed saga orchestrator: forward execution with automatic backward compensation on failure, fluent .add_step() API.

Messaging patterns (v0.2.0):

• outbox.py — transactional outbox for at-least-once delivery: write events in the same DB transaction, relay separately.
• kafka_envelope.py — Kafka-aware envelope: partition key routing, schema version, DLQ routing, producer/consumer roundtrip serialization.
• webhook_handler.py — HMAC-SHA256 webhook verification compatible with GitHub, Stripe, Salesforce, and ServiceNow signature formats.

Change Data Capture (v0.2.0):

• cdc_event.py — typed CDC event envelope: INSERT/UPDATE/DELETE/SNAPSHOT/TRUNCATE, Debezium format parsing, changed-field diff, audit dict.

Ecosystem

See ECOSYSTEM.md for full broker, connector, and framework coverage matrix.

Repository structure

src/integration_automation_patterns/
├── event_envelope.py         # Reliable event transport + retry
├── sync_boundary.py          # Bi-directional SOR sync
├── circuit_breaker.py        # CLOSED/OPEN/HALF_OPEN state machine
├── saga.py                   # Distributed saga orchestrator
├── outbox.py                 # Transactional outbox pattern
├── kafka_envelope.py         # Kafka-aware event envelope
├── webhook_handler.py        # HMAC-SHA256 webhook verification
└── cdc_event.py              # Change Data Capture event types
docs/
├── architecture.md
├── implementation-note-01.md
├── implementation-note-02.md
└── adr/

Published notes

• docs/implementation-note-01.md — event-driven integration reliability
• docs/implementation-note-02.md — idempotency in enterprise event processing

Intended audience

• enterprise architects
• integration engineers
• workflow and automation operators
• platform teams responsible for system-of-record reliability across CRM, ERP, and service platforms

Citing this work

If you use these patterns in your work, see CITATION.cff or use GitHub's "Cite this repository" button above.

---

Part of the enterprise AI patterns trilogy

Library	Focus	Regulation
enterprise-rag-patterns	What to retrieve	FERPA identity-scoped RAG
regulated-ai-governance	What agents may do	FERPA, HIPAA, GLBA policy enforcement
integration-automation-patterns	How data flows	Event-driven enterprise integration