Qudirah Alimi

Machine Learning Engineer with 3+ years of experience designing, building, and deploying production machine learning systems across fintech, humanitarian research, grantmaking, and AI automation. Experienced in developing ML pipelines, training workflows, recommendation systems, predictive models, and LLM-powered applications using Python, scikit-learn, XGBoost, TensorFlow, Docker, Kubernetes, MLflow, and AWS. Built end-to-end ML solutions from business problem definition through data preparation, feature engineering, model evaluation, and deployment, including recommendation engines processing 70,000+ records, credit risk models on large-scale transaction data, and LLM-based decision support systems validated by expert evaluators.

Strong background in production ML engineering, MLOps workflows, model experimentation, NLP, recommender systems, and translating complex technical solutions into measurable business outcomes.

Machine Learning Engineer at Arebak LLC (2023-06 – Present)

• Built an end-to-end grant recommendation engine enabling community-based organizations to discover relevant funding opportunities by developing semantic matching pipelines using TF-IDF and BERT embeddings across 70,000+ grant records, achieving ~80% Precision@5.
• Developed an LLM-based verification agent using OpenAI GPT API, prompt engineering, and custom scoring logic to automate grant evaluation workflows, reducing evaluation time by ~85%, saving ~2 hours per grant cycle, and achieving ~95% agreement with expert evaluators from community-based organizations.
• Engineered and deployed a credit risk ML system for mobile financial services by developing an XGBoost multi-class classification model on large-scale transaction and account data, achieving 81% recall on defaulters at 75% precision on imbalanced datasets.
• Productionized ML workflows by implementing Docker-based deployment, MLflow experiment tracking and model management, GitLab CI/CD integration, and deploying containerized ML services to an existing Kubernetes cluster through GitLab Container Registry.
• Built a predictive disaster-risk model for Heising-Simons Foundation by processing ten years of geospatial data, engineering elevation and water-proximity features using QGIS, and developing an XGBoost model that achieved 70% agreement with observed displacement outcomes.
• Created Tableau reports translating disaster-risk modelling and regression analysis findings into actionable insights for non-technical stakeholders at Heising-Simons Foundation.
• Led internal technical knowledge-sharing sessions and delivered weekly progress reviews to leadership, communicating project progress, technical decisions, and outcomes.

Data Science Instructor at Malhub (2023-01 – 2023-04)

• Designed and delivered a Python and applied data science curriculum covering exploratory data analysis, regression, classification, and model evaluation, improving learner Python proficiency to ~80% and SQL/PostgreSQL competency to ~90% based on assessments.
• Mentored students through end-to-end data projects, helping them apply data analysis workflows and communicate technical findings to non-technical audiences.

Data Analyst at Fezzant (2022-01 – 2022-12)

• Analyzed customer and social media datasets using Python, Pandas, Matplotlib, and Seaborn to identify audience segments and inform targeted content strategies, contributing to a 60% increase in engagement.
• Developed automated web scraping and data preparation pipelines using Python and Scrapy, reducing manual data collection and preparation time by 80%.

Machine Learning Engineer at Cyberodane (2022-09 – 2022-11)

• Built a gradient-boosting classifier for phishing email detection using stylometric features, achieving 87% accuracy and 88% recall.

Bachelor of Science (B.Sc.) in Computer Science – University of Ilorin (2017-09 – 2022-08)