abdifatah.uk
Cardiff, UK · Junior Data Analyst

I turn public and operational data into decisions teams act on.

SQL, Python, and AI-assisted analytics — from LightGBM fare models and supply-chain KPIs to community impact reporting in Butetown. Reproducible pipelines, not one-off spreadsheets.

See selected work → GitHub ↗
−0.91
Price elasticity estimated on NYC fare data
Log-log demand model
17,821
Street-level incidents analysed for Butetown
36 months · data.police.uk
100
SKUs scored across cost, supplier & quality risk
Supply-chain KPI pipeline
01

Selected work

Community analytics · Cardiff

Butetown Community Impact Analysis

A data-driven report on crime, deprivation, and how grassroots sport clubs fill gaps left by structural failures in employment, education, and public safety. Built as Vice-Chair of Tiger Bay Youth Development for committee, parent, and funding conversations.

Street-level crimes by Cardiff area over 36 months
36
API months pulled
87/1,909
WIMD rank (Wales)
260
Crimes / 1,000 people
PythonPandasSQLSQLiteMatplotlibdata.police.uk APIGeoJSON

So what: Butetown's crime rate sits ~38% above the Wales average while ranking in the top ~5% most deprived areas nationally — evidence used to support knife-crime and youth-programme funding bids.

View analysis code ↗
AI / ML · Pricing

NYC Taxi Pricing Analytics

An end-to-end fare-modelling pipeline: OLS regression against official tariffs, a LightGBM predictor, log-log demand elasticity, and an explore-then-exploit online pricing simulation. All steps are importable, testable functions — not a one-off notebook.

−0.91
Price elasticity (ε)
0.51
Demand model R²
$3.02/km
Optimised median price
PythonLightGBMScikit-learnStatsmodelsNumPyShapely

So what: A 10% rise in price per km maps to roughly a 9% fall in distance sold — inelastic but material, giving operators a defensible band to test fares before policy changes.

View analysis code ↗
Operations · KPI analytics

Supply Chain Analytics

A reproducible module turning raw supply-chain records into executive KPIs: revenue and cost structure, a supplier scorecard on lead time and defects, logistics performance by carrier, and quality-risk flags — across 100 multi-channel retail SKUs.

Revenue by product category across the supply-chain dataset
100
SKUs analysed
57%
On-time delivery
2.28%
Avg defect rate
PythonPandasNumPySQLSQLiteMatplotlib

So what: 32 SKUs breach the 3% defect threshold and two suppliers drive most of the quality and lead-time risk — a clear shortlist for procurement to renegotiate.

View analysis code ↗
Community · Excel / BI

Youth Football Participation

Attendance, age-group trends, and behavioural incidents tracked to plan sessions and allocate coaches at a Cardiff youth organisation.

100+
Players tracked
8–16
Age groups

Incidents fell once data review was paired with structured parent and school conversations.

SQL · Databases

SQL Database Capstone

Querying, cleaning, and analysing a relational dataset — joins, conditional aggregates, subqueries, and window functions on messy real-world tables.

IBM
SQL certificate
4-step
Query → clean → analyse

Demonstrates the SQL + Python foundation behind every other project here.

Web analytics · Reporting

Google Analytics Traffic

Sessions, channels, and engagement analysed with the Ask–Prepare–Process–Analyse–Share framework, summarised for non-technical stakeholders.

Channels
Acquisition focus
Report
Stakeholder output

Built during the Google Data Analytics certificate; shows clear data storytelling.

02

What I work with

AI & Machine Learning

  • LightGBM
  • Scikit-learn
  • Feature engineering
  • Model evaluation (RMSE, R²)
  • AI-assisted analysis workflows

Languages & Tools

  • SQL
  • SQLite
  • Python
  • Pandas
  • Excel
  • Git

Visualisation & BI

  • Matplotlib
  • Tableau
  • Google Analytics
  • Stakeholder reporting

Data engineering

  • REST API extraction
  • KPI dashboards
  • Automated chart pipelines
  • SQLite warehousing
03

About

Computer Science graduate (Cardiff Met) with Google, IBM SQL, and Meta data certifications. I build reproducible pipelines in SQL and Python, use gradient-boosted models where they add value, and ship stakeholder-ready visuals — not one-off spreadsheets.

I work with modern AI-assisted development to move faster on cleaning, querying, and documentation while keeping analysis transparent and reproducible. As Vice-Chair at Tiger Bay Youth Development, I connect analytics to real decisions in Butetown — public APIs, SQL, and executive reporting for youth programmes.

Education & certifications

BSc Computer Science (Hons)
Cardiff Metropolitan University
  • CERTGoogle Data Analytics Professional Certificate
  • CERTGoogle: Foundations — Data, Data, Everywhere
  • CERTIBM: Database & SQL for Data Science with Python
  • CERTMeta Database Engineer Professional Certificate
04

Contact

Let's build something with your data.

Open to entry-level data analyst, analytics, and AI-adjacent roles in Cardiff, remote UK, and the public sector. I combine SQL, Python, and practical AI/ML for reports teams actually use.

Emailafm.dvx@gmail.com → Phone07869 731897 → GitHubgithub.com/afmdvx-hue ↗