Manufacturing ERP & WMS Bridge

01 / Business Challenge

• Order and shipment updates sometimes arrived in the wrong order between the warehouse system and the ERP.
• The same warehouse event could post to finance twice, with no reliable way to detect duplicates.
• ERP maintenance and peak shipping windows collided; when something failed, teams needed a safe way to retry without corrupting data.
• When message queues backed up, there was no clear operational picture of what was stuck or why.
• New services had to run in containers and respect existing network boundaries (DMZ, VPN) and security standards.

02 / Our Approach

How we executed this engagement in practice. The phases below describe the delivery rhythm we use across ServiceNow, custom engineering, and mobile programs.

We placed a dedicated .NET 8 integration layer between the warehouse platform and the ERP, using the client’s approved RabbitMQ cluster for messaging. Each message carried stable business identifiers; because delivery can repeat in real systems, we stored deduplication data in SQL Server so finance never saw the same logical event twice. Hangfire ran scheduled checks and dead-letter inspection. A small Blazor operations console gave IT and supply-chain leads queue depth, failures, and controlled replay—with guardrails so retries stay safe.

• Canonical message schemas for order release, pick confirmation, shipment, and inventory adjustment.
• Outbox pattern from WMS-facing APIs to ensure publish reliability.
• Retry policies with exponential backoff and poison-message quarantine.
• Hangfire dashboards restricted to operations roles for controlled reprocessing.
• Health checks and metrics endpoints integrated with existing Prometheus and Grafana.
• Documentation and handover for on-call rotation and ERP cutover windows.