Case studyInformation Systems Analyst
Leading the quality of a WMS — Cafam (2020–2021)
Changing the system that runs a medicine warehouse without stopping dispatch: six months of testing with the largest team I have led.
- WMS
- BI
- VBA
By the numbers
20
people: 14 from Cafam and 6 from Oracle
~6
months of WMS testing
−25%
errors in the inventory record
−50%
data errors with the VBA integrations
Context
Cafam was implementing a WMS across its logistics operation: a critical system change where every data error is paid for in the warehouse.
The challenge
Assure the quality of the implementation while coordinating the largest testing team I've led, without slowing the operation.
How I did it
First the process, then the system
Before the first test I modeled the distribution center in BPMN with Bizagi —receiving, storage, picking and dispatch— and simulated dispatch in FlexSim to know where a new WMS rule could create a queue that did not exist before.
A mixed team of twenty
I led 14 people from Cafam and 6 from Oracle through about six months of Oracle WMS Cloud testing. My job was to translate: an inventory complaint became a reproducible scenario, with inputs, steps, an expected result and evidence.
Instrumenting the implementation
I built the control BI for the project itself for more than 15 users, from directors to distribution-center coordinators: coverage, defects, recurrences and retests visible every day, not a feeling of progress.
Integrations with an owner and a limit
I developed VBA integrations with the WMS to replace manual loads. Each peripheral piece knew who maintained it, what it depended on and when it would be retired: without those three conditions, a bridge becomes a parallel system.
The 80-20 reconciliation in SQL
At the start, none of the main medicines matched between the WMS and the source system. I reconciled them in SQL by volume priority, and every query that found a difference stayed as a permanent control.
Impact
- −25% errors and +15% operational efficiency across the implementation.
- +50% precision in operational control, adopted by 15+ users.
- +15% automation and −50% data errors with the VBA integrations.
- +20% data quality via SQL.
Modernizing does not consist of replacing everything at once, but of designing a transition that preserves continuity while progressively reducing complexity.