MSc Artificial Intelligence (Distinction) graduate building data platforms, practical AI and decision-support systems that get used.
I’m interested in AI systems that move beyond notebooks into usable tools: medical imaging workflows, explainable ML, route-aware decision support, multi-agent simulation, and human-in-the-loop operational systems.
Production-shaped medallion lakehouse over the National Grid ESO Carbon Intensity API and Open-Meteo: append-only bronze with SHA-256 manifests, pandera-validated Parquet silver with a quarantine, DuckDB views, 20 dbt models with 61 tests, Postgres serving (COPY + upsert), a FastAPI read API, Airflow DAGs with catchup and a quality gate, Docker Compose, GitHub Actions CI and a nightly live run. Backfilled 90 days and quantified ESO's day-ahead forecast error and weather-to-generation correlations by region.
Research prototype for automated L3 skeletal muscle analysis from abdominal CT scans. Includes DICOM/NIfTI processing, L3 slice selection, TotalSegmentator baseline masks, PyTorch U-Net segmentation, CSA/SMRA/SMI metric calculation, aggregate evaluation plots, and a Streamlit review interface.
End-to-end tabular machine learning workflow for used-car price prediction, including preprocessing, ensemble models, model comparison, feature importance, and explainability.
Comparison of BFS, DFS, UCS, and A* search algorithms on real OpenStreetMap road-network data, with benchmark outputs and runtime analysis.
Computer vision project for vehicle detection and classification using YOLOv8, annotated image data, validation metrics, and sample predictions.
Drone-based multi-agent search simulation for oil-spill detection under uncertainty, modelling search coverage, sensing radius, environmental dynamics, and agent coordination.
C# data-structures project implementing a binary-search-tree text indexer with search and lookup functionality.
Data Engineering: Parquet, DuckDB, dbt, Airflow, Postgres, pandera, Docker, GitHub Actions Machine Learning & AI: PyTorch, scikit-learn, YOLOv8, OpenCV, SHAP, Gemini API Data & Scientific Computing: pandas, NumPy, matplotlib, Jupyter Medical Imaging: DICOM, NIfTI, SimpleITK, segmentation workflows Software & Tools: Python, C#, SQL, Streamlit, Git, GitHub Current Direction: explainable AI, decision-support systems, route-aware optimisation, and human-in-the-loop tools
At Time Specialist Support I built the organisation's first data and automation systems: an explainable staff/session allocation engine (constraints, scoring, travel feasibility, human review), a Gemini-powered safeguarding monitor, an OCR-to-JSON profile pipeline and return-to-work automation. Open to data engineering and AI engineering roles in Manchester and across the UK.
- GitHub: @Isaac-Ashamu
- LinkedIn: www.linkedin.com/in/inioluwa-ashamu