Adeniyi Lawal

Manchester, UK

Data analyst turning complex data into clear, actionable decisions.

Specialising in healthcare and operational analytics, fairness auditing and predictive modelling.

Get in touch See my projects
10 yearsdata analytics experience
MScBusiness Analytics & Technology
9AWS certifications
ToolsPython, R, SQL, Power BI, AWS
Adeniyi Lawal in a dark suit against a city backdrop

Selected work

Three projects. Each starts with a question, uses a method, and ends with a result.

Algorithmic fairness · Python · Streamlit

FairHire Audit Tool

Found bias in machine-learning models used for recruitment, then built a live web tool that audits a candidate dataset for it.

  • Random Forest gender Disparate Impact Ratio rose from 0.767 to 0.862 after mitigation, above the 0.80 practical benchmark.
  • SHAP analysis flagged a proxy variable (RecruitmentStrategy_1) that could discriminate indirectly.
  • Compares five fairness metrics across three classifiers and two datasets.

PythonSHAPAIF360FairlearnStreamlit

Live appCode on GitHub

Bar chart of Disparate Impact Ratio before and after mitigation for five groups and models

Epidemiology · Survival analysis · R

Coronary heart disease risk factors

Identified which measured factors are linked to heart disease over 24 years, using the Framingham Heart Study teaching dataset.

  • Analysed 4,013 participants with complete baseline data.
  • Men had a hazard ratio of 2.26 compared with women. Diabetes was 1.57 and current smoking 1.20.
  • Kaplan-Meier curves show men reach heart disease earlier than women (log-rank p < 0.001).

RCox regressionKaplan-MeierLASSOROC / AUC

Code on GitHub

Teaching dataset, used for analysis only. Age and prior hypertension break the proportional-hazards assumption, so their hazard ratios are averages over follow-up.

Forest plot of hazard ratios for heart disease risk factorsKaplan-Meier curves of heart-disease-free survival by sex

Operational analytics · R · NHS England data

COVID-19 bed occupancy across NHS England

Cleaned and compared regional COVID-19 bed occupancy data to show where the load fell between April and June 2026.

  • London carried the highest average daily COVID-19 bed occupancy of the six regions.
  • Built a regional comparison and a trust-level view to support resource planning.

Rtidyverseggplot2Data cleaning

Code on GitHub

Bar chart of average daily COVID-19 bed occupancy by NHS regionBar chart comparing one hospital trust with the London regional average

Skills and credentials

What I use day to day, and what has been verified.

Machine learning and analysis

Python (pandas, numpy, scikit-learn, XGBoost, lifelines), R (survival, glmnet, caret, pROC). Classifiers: logistic regression, random forest, XGBoost. Cox regression, Kaplan-Meier, LASSO, ROC/AUC.

Fairness and explainability

SHAP explainability, AIF360 fairness metrics, Fairlearn bias mitigation (ExponentiatedGradient, ThresholdOptimizer), proxy-variable detection.

Data and dashboards

SQL, Power BI, Streamlit for live apps, Microsoft 365 (Excel), data integration and validation, and reporting that non-technical people can act on.

Cloud and AI

AWS data engineering, machine learning, NLP and generative AI. Full list in Certifications below.

Education

MSc Business Analytics and Technology, University of Greater Manchester, UK.
BSc Accounting and Finance (Second Class Honours), Ajayi Crowther University, Nigeria.
Data analytics training, Codar Institute (Power BI, Excel).

Adeniyi Lawal walking through a city street

About me

Read my story

I started my career in accounting and finance. I moved into sales in a retention and conversion team, where I saw first-hand which customers stayed, which left, and why. I then trained as a teacher, and later progressed to supervisor, team leader and manager, taking responsibility for people, targets and decisions.

Across every one of those roles I kept running into the same problem: the answers were already in the figures, but nobody had the time or the tools to pull them out. Learning to do that myself is what led me into analytics, and the projects on this page are the result.

I like work where the data is imperfect and the stakes are real: recruitment decisions, hospital capacity, patient risk. I check my results, say where they break, and explain them in plain language.

I am flexible on shift patterns.

Dashboards and analysis

Business and operational dashboards built in Power BI and Excel, with clear answers to the questions behind them.

1 / 4

Insurance · Power BI · Excel

Insurance claims analysis

Analysed around 38K policyholders to see how claim frequency and claim amount differ by age, gender, marital status, education, coverage zone and vehicle make.

  • Claim frequency is lowest in highly rural zones and highest in suburban zones.
  • Built a question-and-answer dashboard so each finding is traceable to the data.
  • Recommendations on premium pricing and targeted campaigns.

Power BIExcelData cleaningSegmentation

Power BI dashboard of insurance claim frequency and amount by demographic groupPower BI dashboard of claim amount and frequency by vehicle make and model

Retail · Power BI

Sales, product and customer analysis

Three linked dashboards covering customers, products and time, so a manager can move from a headline number to the detail behind it.

  • Customer view: 792 distinct customers and 9,716 orders, with revenue split by segment, region and shipment type.
  • Time view: sales and profit by quarter, month and day.

Power BIDAXInteractive filters

Power BI customer analysis dashboard with revenue by segment and regionPower BI time series dashboard with sales and profit by quarter and month

Sales · Excel · Power BI

Coffee and gadget sales dashboards

Two sales dashboards: a coffee range dashboard in Excel and a gadget sales dashboard in Power BI.

  • Coffee: sales and budget compared by product, region and year, with profit by territory.
  • Gadgets: about 34M in revenue across 19 products, with revenue by city and product.

ExcelPower BIKPI cards

Excel dashboard of coffee sales, budget and profitPower BI gadget sales dashboard with revenue by product and city

HR analytics · Power BI

HR attrition analysis

Looked at 1,470 employees to understand who leaves and where attrition is concentrated.

  • 237 employees attrited, an attrition rate of 16.12%.
  • Breakdowns by job role, age band, years of service and education, to target retention effort.

Power BIHR analyticsRetention

Power BI HR attrition dashboard with job role, age and service breakdowns

Practice and growth

How I keep learning between projects, and what I do with the feedback I get.

AI tools

How I use AI in my work

I use AI assistants to speed up drafting, coding and checking. I verify every output against the source data before I rely on it.

ClaudeDeepSeekMicrosoft 365n8nGoogle ColabStreamlit

Hobby project

LinkedIn banner design

A personal design project: a dark, gold-accented banner that carries my skills and links, built to match the look of this portfolio.

DesignBrand consistency

Mentorship

Feedback, then the next step

I share my work with a mentor for corrections before I move on. The feedback so far has shaped this page: clearer project write-ups, and showing practice work alongside client-facing projects.

Certifications

Verified credentials, grouped by focus.

Core data and analytics 2
  • Google Data Analytics Professional CertificateGoogle / Coursera · 2022
  • Advanced Excel CertificationCoursera · 2022
Cloud and AI (AWS) 9
  • AWS Data Engineering2026
  • AWS Machine Learning Foundations2026
  • AWS ML for Natural Language Processing2026
  • AWS Generative AI Foundations2026
  • AWS Cloud Architecting2026
  • AWS Cloud Foundations2026
  • AWS Cloud Security Foundations2026
  • AWS Cloud Operations2026
  • AWS Data Center Technician2026

Let's talk

Open to data analyst roles, and to work in healthcare and operational analytics.

Personal email

princelawaladeniyi@gmail.com Best for recruiters and collaborators

Phone

+44 7584 559996 UK mobile
Email me LinkedIn GitHub