The warehouse was picking hundreds of orders a day on paper. The ERP-supported shipping options were inadequate, and the business had been told that connecting a better cloud shipping platform was not possible. The constraint was not a lack of software. It was the gap between the software and the real operation.
The approach
I worked with operations to document the workflow and its edge cases, then reverse-engineered the ERP's picking and shipping behavior through SQL traces, queries, commands, and stored procedures.
What I built
I personally scoped, architected, wrote, deployed, and supported two connected production systems:
- A tablet-based picking application using shelf-level QR scanning
- FDA-regulated lot allocation and traceability
- Custom middleware between the cloud shipping platform and the self-hosted ERP
- Bidirectional API and database integration
- Automated order confirmation and document rendering
- Dynamic rate shopping against customer delivery requirements
The picking application used Flutter, Node.js, SQL Server, Android tablets, and Zebra hardware. The shipping layer bridged a modern cloud service with a legacy system that had no supported integration path.
The outcome
The combined system replaced paper workflows, improved traceability, automated repetitive shipping work, and made a previously unavailable cloud platform usable inside the existing ERP environment. It also created meaningful operational savings without forcing a disruptive ERP replacement.
Why this is signature work
This project captures the unglamorous, high-value work I enjoy most: learning an operation in detail, reverse-engineering the systems underneath it, and building the missing layer myself. It joined software, hardware, warehouse process, legacy ERP behavior, regulatory traceability, and vendor coordination into one production workflow.