🏭 Supply Chain · SAP · Power BI · SQL

Real-Time Operational Performance Dashboard & ERP Data Fix

Built a unified live dashboard for Jabil — one of the world's largest manufacturing companies — consolidating 3 separate business units and fixing the root-cause data quality issues in their ERP system.

3
Business units unified
1 day
Manual work eliminated
Live
vs weekly reports
Fixed
ERP data mismatches
The Problem

Jabil had 3 business units each tracking their own numbers in separate Excel files. Every week, someone spent a full day manually collecting those files, combining them, and building a report for leadership. The numbers were always a week old by the time anyone saw them.

On top of that, the ERP system had a data quality problem that nobody had properly traced. The inventory system would show 500 units in stock, but the financial system would show costs for 480 units. Nobody knew which number was right — so both were wrong.

Leadership was making operational and budget decisions without being able to trust the data in front of them.

The Goal

Build one central dashboard pulling live data from all 3 business units automatically — so leadership had one clean set of numbers without waiting. And separately, trace the ERP data discrepancies to their root cause and fix them at the source, not patch them in reports.

My Role

I was the data analyst on this. I built the Power BI dashboards, wrote the SQL to pull from SAP, investigated the ERP discrepancies by comparing numbers across tables, worked with IT and the ERP vendor to get them fixed, wrote the functional requirements for the changes, and ran UAT to confirm the fixes actually worked before anything went live.

The Team
💻
IT Team
Had access to the SAP backend — worked with them to trace discrepancies and implement fixes
🔧
ERP Vendor
External team managing the SAP software — helped resolve the technical configuration issues
💰
Finance Manager
Defined what metrics leadership needed and validated that dashboard outputs matched expectations
⚙️
Operations Managers (×3)
One per business unit — explained how their data was generated and what the numbers meant in context
Tools & What I Used Them For
SAP S/4HANA ERP System
Pulled operational data from Jabil's ERP system and used it to investigate where numbers were going wrong — comparing inventory counts, financial postings, and order data across modules to find the discrepancies.
Power BI Dashboards
Built the live central dashboard showing inventory levels, service rates (on-time order fulfillment), throughput, and cost variance across all 3 business units. Leadership stopped waiting on weekly Excel compilations.
SQL Data Investigation
Wrote queries comparing numbers across different SAP tables to find mismatches. Like checking if the inventory count in one table matched the financial cost count in another — and tracing which table was wrong and why.
Excel Forecasting
Built demand forecasting models using historical production and supply data. The models predicted how much material each line would need so supply and demand mismatches on the manufacturing floor could be caught earlier.
UAT — User Acceptance Testing QA
When IT made changes to fix the ERP issues, I wrote the test cases and ran through each one to confirm the fix worked correctly before anything went live. Found two issues in testing that would have caused new problems if they'd shipped.
Outcomes
📊
Leadership had live numbers. Real-time dashboard replaced a weekly manually compiled report. One day of work eliminated every week.
✅
ERP data mismatches traced and fixed. Inventory and financial figures started agreeing with each other. Reports became trustworthy again.
📈
Supply-demand mismatches reduced. The forecasting model gave production lines better visibility into upcoming material needs, reducing last-minute shortages and overstock.
🔒
ERP changes went live without issues. Two problems caught in UAT before go-live — both would have created downstream data errors if they'd shipped.
What a Typical Day Looked Like
Morning
Check Power BI dashboards for unusual numbers. If something looked off, pull the data in SQL to find out why before the operations team's day started.
Mid-Morning
SQL work — either pulling reports for leadership, investigating a specific data discrepancy, or comparing tables to find where numbers were diverging.
Afternoon
Meetings with IT, operations managers, or the ERP vendor. Sharing findings, getting context on what the data should look like, reviewing fix proposals.
End of Day
Update forecasting models, finalize the weekly leadership report, or run through UAT test cases if the IT team had pushed a change.
← CHFA Project Next Project: YMCA →