About

I turn scattered numbers into decisions.

I'm Ammar Ullah Khan, a Microsoft Certified Power BI Data Analyst (PL-300). I build the reporting layer that turns the Monday leadership meeting from ninety minutes of debate into ten minutes of decisions.

Right now I'm embedded with Foxtons, London's largest estate agency, reporting to the Head of Business Analytics. I've delivered 24 production dashboards across their two highest-revenue departments, tracking 40+ KPIs that senior leadership uses to make budget and staffing calls.

Before Foxtons I built analytics from zero for clients across Austria, Saudi Arabia, the United States and Africa — including proof-of-concept work for IKEA, a fleet telematics platform for PowerTrack Africa, and CFO-ready reporting for a PIF-backed aviation operator.

What you get isn't a dashboard. It's the specific number your team argues about, on a screen, refreshed automatically, that nobody has to rebuild on Monday morning.

Why my dashboards get used

Most BI projects die the same way. Someone builds a beautiful dashboard. It's technically correct. Nobody uses it, because it answers questions nobody was asking, and the numbers don't reconcile with what the finance team already has in Excel.

I've spent my career learning why that happens and how to prevent it. The method is unglamorous: verify before you build. Understand what decision the report is supposed to change. Reconcile every figure against the source system before anyone sees a chart. Then, and only then, design.

That's it. That's why my dashboards get used.

How I work

Diagnose

The first conversation is about decisions, not data. What are you arguing about in meetings? What gets rebuilt by hand every week? What number does nobody trust?

Verify

Before I build anything, I reconcile your figures against your source systems. Almost every engagement surfaces at least one discrepancy nobody knew existed. This step is where trust gets earned, and it's the step most people skip.

Build

Power BI, with a proper semantic model underneath so the report doesn't collapse the first time someone asks a question you didn't anticipate. Row-Level Security, scheduled refresh, version control, documentation.

Hand over

Data dictionary, refresh schedule, and a walkthrough so your team owns it. I don't build dependency.

What I work with

BI & Visualisation
Power BI DAX Power Query Row-Level Security Scheduled Refresh Tableau AWS QuickSight Looker Studio
Microsoft Power Platform
Power Automate Power Apps SharePoint Teams Embedding Microsoft Fabric
Data Engineering & ETL
SQL Python Apache Airflow PySpark Azure Data Factory Azure Synapse
Cloud & Databases
AWS Redshift Amazon RDS Azure BLOB PostgreSQL MySQL
Other
MLflow Scikit-learn Excel Git & GitHub Data Modelling

Let's find out whether this fits.

Book a 20-minute call. I'll ask what decision you're trying to make and tell you honestly whether I'm the right person to help. If I'm not, I'll say so, and I'll usually know who is.