Data Scientist • AI/ML Enginneer • OSS Contributor • NLP • Computer Vision • LLMs • MLOps • Computer Science
Hi — I’m Charles Akinnurun, currently working as a Data Scientist (VIRTUAL EXPERIENCE PARTICIPANT) at British Airways.
As a Data Scientist, AI/ML Engineer and an OSS Contributor that specializes in data science, machine learning, deep learning, natural language processing (NLP), computer vision, and artificial intelligence, with a proven track record of building scalable, production-ready AI/ML systems that deliver measurable business and technical impact.
I have led and executed the end-to-end machine learning lifecycle across 140+ ML projects, spanning data collection, preprocessing, feature engineering, exploratory data analysis (EDA), model development, evaluation, deployment, monitoring, and optimization using Python (Scikit-learn, PyTorch, TensorFlow, Keras, Pandas, NumPy, Seaborn, Matplotlib), R, and SQL. These projects processed datasets ranging from 100K to 10M+ records, supporting batch and real-time inference systems serving up to 5,000+ requests per minute.
Through systematic experimentation and advanced hyperparameter tuning using XGBoost, LightGBM, and CatBoost, I improved model performance by 38%–45%, reduced error rates by 27%–40%, increased AUC scores from 0.71 to 0.89, and reduced inference latency by 30%, enabling near real-time decision-making in production environments.
My work spans regression, classification, clustering, association rule mining, dimensionality reduction, time-series forecasting, and real-time inference systems. These models have:
- Improved regression model performance with R² increased by 0.31, RMSE reduced by 22%, MAE reduced by 19%, and MAPE reduced by 17%.
- Increased prediction accuracy by up to 42%, improving F1-Score by 0.29, Recall by 18%, and ROC-AUC by 0.21.
- Increased Precision by 24% while reducing False Positive Rate (FPR) by 21%.
- Boosted ROC-AUC by 0.21 and PR-AUC by 0.19, strengthening class discrimination performance.
- Reduced Davies–Bouldin Index by 22%, demonstrating better cluster compactness and separation.
- Increased Calinski–Harabasz Index by 31%, reflecting improved cluster structure and variance ratio.
I bring a strong foundation in computer science, applied mathematics, statistics, probability, data structures and algorithms, enabling me to design robust, scalable, and interpretable systems handling multi-terabyte datasets and distributed workloads.
I design and manage scalable database systems supporting AI/ML workflows using PostgreSQL, MySQL, and SQLite, building optimized schemas and high-performance queries. I have developed and maintained 50+ database-driven pipelines, reducing query latency by 35–60%, improving data processing throughput by 40–55%, and lowering infrastructure costs by 20% through indexing, partitioning, caching, and query optimization.
By implementing robust ETL/ELT pipelines, automated validation checks, and data quality monitoring systems, I increased data reliability by 30–45%, reduced pipeline failures by 50%, and ensured 99%+ data consistency across production environments.
I have implemented automated retraining workflows that reduced manual intervention by 70%, established monitoring systems detecting data drift and model degradation within 24 hours, and reduced model downtime by 60%, maintaining SLA compliance above 95%.
In addition to hands-on development, I actively contribute to open-source projects, collaborate with cross-functional teams (engineering, product, analytics) across 25+ projects, accelerate experimentation cycles by 35%, and regularly transform research prototypes into production-ready systems within 2–4 weeks, reducing time-to-deployment by 40%.
“Code. Learn. Build. Repeat.”