Open to collaborations and opportunities

AI ResearcherData ScientistStatistician

Chinedu
Uzim.

PhD researcher in causal AI at Leeds Beckett University. I develop domain-agnostic methods that uncover cause and effect in messy, real-world data, using food security systems as a use case, and I build AI products that people use.

PhD researcher
Leeds Beckett University
MSc, Distinction
Data Analytics and Technologies
AI products shipped
SmartSkan and UziIQ
Chinedu Uzim
Based in Leeds, UKOpen to UK and international roles

01About

Asking why, not only what next.

I'm Chinedu Matthew Uzim, an AI researcher, data scientist and statistician based in Leeds. Most models are built to predict what will happen next. I'm interested in the harder question underneath: why it happens, and what would change if someone acted on it.

Alongside my PhD at Leeds Beckett University, I design and ship AI products. I built SmartSkan, Skannr's symptom-to-scan recommendation system, and UziIQ, a live review-analysis tool.

Earlier in my career, I worked with national survey, immunisation and electricity data at the National Bureau of Statistics, the National Primary Health Care Development Agency and Eko Electricity Distribution Company in Nigeria.

Education

2025 – present
PhD, Computer Science and Engineering
Leeds Beckett University, UK
2024 – 2025
MSc Data Analytics and Technologies, Distinction
University of Greater Manchester, UK
2022
BSc Statistics, Upper Division
Alex Ekwueme Federal University Ndufu-Alike Ikwo, Nigeria

02Research

From prediction to cause and effect.

Doctoral research · Leeds Beckett University · 2025 – present

Causal AI for decisions in the real world

Most AI systems are good at prediction but say little about cause and effect, which is what decision-makers need when they ask what would happen if they intervened.

My PhD develops causal AI methods, with a focus on causal discovery: learning cause-and-effect structure from observational data that is short, noisy, incomplete or changing over time. The methods are designed to be domain-agnostic, with food security systems as the use case for validation, where stronger causal evidence could support earlier and better-targeted interventions.

Themes

  • Causal discovery from time series
  • Reliable methods under distribution shift and missing data
  • Moving from forecasts to interventions
Journal article Co-author May 2026

Development of a Personalized Geographic Information System (PGIS) for Precision Agriculture to Optimise Rice Production and Increase Its Yield in Ebonyi State, Nigeria

Nwogbu Peter Chinedu; Ezza Fon Scholars; Chinedu Matthew Uzim; Nwigwe Simon; Ngwuta Amuche Daniel; Ofobuike Oketa Nwali

International Journal of Innovative Science and Research Technology (IJISRT), Vol. 11, Issue 5, pp. 154–174. DOI: 10.38124/ijisrt/26may211

The study develops a personalised GIS framework that combines Sentinel-1 SAR remote sensing, geospatial mapping and agricultural statistics to give site-specific recommendations for rice production in Ebonyi State, Nigeria, addressing soil fertility decline, climate variability and inefficient use of resources.

Acknowledged contributions

Credited in the paper, not as an author

Acknowledged for data analysis Conference paper September 2025

The Reach of Digital Games and Their Potential as Global Communication Tools

Yau, J., Harris, R., Hollins, P., Garg, A., Hewage, P., Ower, J., Schuur, J., Griffiths, D., Watson, P., Iwendi, C. and Drachsler, H.

Proceedings of the 19th European Conference on Games Based Learning (ECGBL), Vol. 19, No. 2, pp. 910–917. DOI: 10.34190/ecgbl.19.2.3874

My contribution: I supported the data analysis for Play2Act, a case study run with the United Nations Development Programme within the EU and UKRI-funded GREAT project. Short climate surveys placed inside popular mobile games reached almost one million players and produced about 181,000 completed responses.

03Projects

Selected work.

Medical AI · Skannr

SmartSkan

Skannr's symptom-to-scan recommendation system. Structured medical rules give reliable recommendations for common symptoms, while Google Gemini handles complex cases. It connects to Skannr's provider network, so patients see real scan options with prices and booking.

Python · FastAPI · Google Gemini · React / Vite

AI product · SaaS

UziIQ

A live web tool that analyses customer reviews with a large language model and turns them into clear, usable insight. I designed, built and launched it end to end.

Node.js · Express · Claude API · Railway

Machine learning · MSc

House Price Predictor

Compared XGBoost, LSTM, SVR and decision-tree models for house price prediction, with tuning improving RMSE by about 15%. The tuned XGBoost model powers a public Shiny app that gives real-time predictions.

R · caret · XGBoost · Shiny

Data science · Web scraping

UK Car Market Analysis

Scraped and analysed more than 5,000 used-car listings from The AA and Cinch to explore how price, mileage, transmission and fuel type vary across the UK second-hand market.

R · rvest · httr · tidyverse · Power BI

04Press

In the media.

Skannr, In the Spotlight

Skannr interviewed me about building SmartSkan: balancing AI innovation with patient safety, and connecting people directly to real scan providers.

Read the full feature
I built a hybrid system combining structured medical rules with Google's Gemini AI. This approach delivers reliable recommendations for common symptoms while handling complex cases intelligently. The integration with Skannr's provider network means patients see real options with pricing and booking instantly.
Chinedu Uzim, interviewed by Skannr

05Skills

Toolkit.

Causal and statistical modelling

Causal inference and discovery, regression, time series analysis, forecasting, correlation analysis

Machine learning

XGBoost, LSTM, SVR, decision trees, deep learning, NLP, model tuning and evaluation

AI product development

FastAPI, Node.js and Express, React and Vite, Gemini and Claude APIs, hybrid rule-based systems

Data and analytics

Python (pandas, SciPy), R (tidyverse, caret, Shiny), SQL, Power BI, Tableau, SPSS, Excel

Geospatial

Sentinel-1 SAR, GIS, spatial data integration, precision agriculture

Cloud and deployment

AWS Glue and Athena, ETL pipelines, Git and GitHub, GitHub Pages, Railway

06Contact

Let's work
together.

Open to research collaborations, academic partnerships, data science and AI roles, and speaking opportunities.

chineduuzim1@gmail.com