Research & Writing
Journal Articles
Peer-reviewed applied machine learning research, co-authored and submitted to Computers & Industrial Engineering (Elsevier). Author names link to Google Scholar; a title links to the published article where one exists.
-
Blockchain used to make energy exchange records auditable, with machine-learning anomaly detection over the consumption data on top; evaluated against the Ausgrid smart-meter dataset.
-
Dual verification-based multimodal approach for reliable pothole and speed bump detection on embedded system
Road-surface hazard detection running on an embedded target, where two sensing modalities have to agree before a detection is reported — the check that keeps a single noisy channel from raising it alone.
Technical Articles
Hands-on engineering write-ups, self-published on Medium. These are not peer reviewed and make no claim to be; they are working notes on problems solved in full, published so the next person does not have to solve them twice. View and read counts are Medium’s own, as of August 2026.
-
Simplify Your Log Management: Configuring DataDog Integration with Log4j2 In Mulesoft
Ships logs without installing the DataDog agent: a Log4j2 HTTP appender POSTs JSONLayout-formatted events straight to the Send Logs API, behind an async root logger rather than a blocking one.
-
Developing a Custom YOLOv8 Model for Number Detection on Meters Using FastAPI and Gradio
Carries one annotated dataset the whole way through — Roboflow labelling, 100 epochs of distributed training on Kaggle, then the same weights behind two surfaces: a containerised FastAPI service and a Gradio space on Hugging Face.