IT & Business Intelligence Lead. I build and run the data platforms behind sales and distribution — ingestion, warehouse, dashboards, and the operations that keep them alive. Erbil, Iraq.
Nineteen years in IT. Networks first, then SAP, then data, which is where I stayed. I still run all three.
Telecom engineering is where I came in, and the network never left. Enterprise routing and switching on Cisco, firewalls and security policy on Palo Alto. It is the layer nobody notices until it stops, and it is still mine.
Then SAP, learned in production and across modules, because the questions I was being asked lived inside the ERP. It turned out to be the bridge: infrastructure on one side, business data on the other, and after a few years you can read both.
Data is the middle of the work now. I design, build and operate end-to-end BI platforms for sales and distribution across six regions. Distributor files land every morning; automated ETL in Python and SQL Server ingests them, data-quality checks decide which numbers are allowed through, and the dashboards people open — on a laptop or a phone — are built from what survived. Executive reports send themselves. Newer work runs on Microsoft Fabric.
One person owns that whole line: ingestion, warehouse, semantic model, dashboards, operations. The last word is the one that gets underestimated. A dashboard is easy to build once; the work is having it still be right months later, when a distributor quietly changes a file format and tells nobody.
The seat is the real advantage. Most people in this trade stand in one place: a sales manager reads reports someone else built, an analyst builds reports about a business someone else runs. I sit where the two meet, and the view is different there. Service-level KPIs from the fundamentals to the advanced tiers — coverage, distribution, customer health, missing customers, offtake and expiry risk — built as more than fifteen analytical engines and counting, across six regions, on both models of the trade: distributor-led markets and direct distribution.
Finance is the other half. Ten years as an SAP finance key user runs alongside all of it — profit and loss, costing, mix. It changes what a number means. Top line and bottom line are not a sequence; they move together, and a sales push that quietly erodes margin is not growth. Finding the mix that moves both at once is the actual job, and you only see it when the same person reads the sales figures and the P&L.
The newest thread is AI-assisted engineering — workflows where agents do real work, writing code, moving data, drafting the documentation nobody wants to write, inside a discipline that checks them. An agent nobody verifies is just a confident stranger. kibsu is the public piece of that thread.
Work
-
Sales & distribution dashboard
A working demo of the kind of thing I build daily: KPI tiles, monthly trend, top movers, outlet coverage — in an executive view, an analyst view with drill-down, and a phone view. Entirely synthetic data, hand-drawn SVG, no chart library and no external requests. - The same invented numbers read through the P&L. Push volume on one product with a promotion and watch revenue and gross profit move in opposite directions. The verdict is a sentence, not a colour.
- Reads the instructions you have written for coding agents, alongside your git history, and reports which of them anyone could actually verify were followed. Python, no dependencies, MIT.
Writing
Your agent’s instructions are promises nobody checks. I counted.