AI and Machine Learning Engineer

I fine-tuned an AfriBERTa transformer model using a labelled dataset of 219 Hausa-language examples. The model classifies text across six humanitarian crisis categories: conflict, displacement, food insecurity, disease outbreak, flood, and no crisis.

I owned the end-to-end machine learning pipeline, including data collection and labelling, GPU-based training in Google Colab, evaluation and iteration, Gradio interface development, and deployment to Hugging Face Spaces.

I built and evaluated a supervised classification model that predicts crop type from soil metrics using Python, scikit-learn, and pandas.

I implemented train-test splitting, model selection, hyperparameter tuning, and accuracy evaluation across multiple classifiers.

I am building an automated news scraping and NLP classification pipeline that ingests Hausa and English news daily and classifies crisis incidents using a deployed model.

The architecture includes an RSS scraper, NLP classifier, JSON database, Gradio dashboard, and GitHub Actions for daily automation. The project focuses on end-to-end ML system design and automated inference pipelines.

I participated in the full software development lifecycle, including design, testing, release, and post-launch support within an agile sprint framework.

I collaborated with engineering and product teams to gather requirements, deliver iterative builds against sprint milestones, produce technical documentation, and support stakeholder requirement validation.

I managed research databases and prepared datasets for analysis and evidence generation.

I implemented data cleaning and structuring pipelines to improve the usability and organization of research data.

I completed software engineering training covering Python, algorithms, data structures, system design, and collaborative development practices.

The programme strengthened my foundation in software development and team-based engineering workflows.

NYSC completed.