Regional e-commerce operator, key developer gone, inventory accuracy collapsing.

Situation

A regional e-commerce operator running fulfillment out of several warehouses. The internal inventory dashboard and dispatch tool had been built over multiple years by a single in-house developer. That developer had left half a year earlier. The remaining team could run the system but could not modify it. Errors were growing, and the business was making daily operational decisions on inventory data that was a day or more stale.

What was stuck

No active development. No documentation. No one who understood the codebase. Multiple failed attempts to hand the code off to an offshore team had each stalled in the first two weeks because leads couldn't get a working local environment up. The ops team had started maintaining a parallel inventory spreadsheet just to stay operational.

What we changed

We started with a Stabilization Review. Fixed fee, two weeks. The review identified the two most critical data pipelines (inventory sync and dispatch) and mapped the system architecture in a written deliverable. We took over the repo, documented the existing system inside two weeks, and patched the inventory staleness issue in the first week of the retainer by rebuilding the sync process and adding basic observability. Over the next several months on retainer we rewrote the dispatch module cleanly while keeping the legacy module running in parallel until cutover.

Result

Inventory data accuracy improved from a day or more stale to near-real-time against warehouse events. Dispatch error rate dropped materially. The ops team retired the parallel spreadsheet. The client retained Kingbird for additional feature work and warehouse integration projects.

Stack

• Python (Django)
• PostgreSQL
• Redis
• Shopify Admin API
• Warehouse management integrations