Data Analyst & Analytics Engineer

Ridwanllah Husain

I build the data layer underneath a business, then the reporting that runs on it. Most of my work comes down to finding where value quietly goes missing, and what it would cost to stop it.

View Projects → Let's Talk Download CV

Available for new projects

About

I turn messy data into numbers a team can actually decide on.

I'm a data analyst and analytics engineer, and I work the whole way through a problem: getting data out of the systems it's stuck in, shaping it into something you can trust, then building the reports and dashboards a team checks every week. Most of the value sits in that middle part, where a number either survives being questioned or it doesn't.

I studied electrical engineering, so I usually want the reason behind a number rather than just the correlation, and I'll tell you where an analysis stops being reliable before anyone builds a decision on it. The pipelines matter for the same reason: scheduled ETL, a warehouse designed in SQL, loads that are safe to rerun, and checks that catch bad data before it reaches a report.

Available now for full-time analytics roles and freelance work: dashboard builds, one-off analyses, predictive models, and pipeline setup for teams that don't need a full-time data hire yet.

See how that looks in practice →

Services

What I Build

Exploratory work, segmentation, cohort breakdowns and the recurring reports a team runs on. You get the numbers, the reasoning behind them, and a straight answer on what they do and don't tell you.

Interactive dashboards in Power BI, plus lighter app interfaces where a full dashboard would be overkill. Built around the questions people already ask, so they can find answers without coming through me.

Data modelled properly: clean tables, a sensible grain, and metric definitions written down. This is the layer that keeps every report consistent instead of each one quietly telling a different story.

Scheduled pipelines with retries, backfills and quality checks built in. When something breaks you hear about it that day, rather than finding out from a wrong number a week later.

Classification, scoring and forecasting models, tested on data the model hasn't seen and calibrated so the probabilities mean something. Handed over with a plain read on how well it performs and where it falls down.

Selected work