Network Intrusion Detection
Binary and 5-way network attack classification on the real NSL-KDD benchmark, with transparent analysis of where accuracy metrics mislead.
Data engineering, analytics and applied-ML work — each one built end to end.
Binary and 5-way network attack classification on the real NSL-KDD benchmark, with transparent analysis of where accuracy metrics mislead.
A production-style ETL pipeline: raw retail exports flow through a validated data-quality gate into a star-schema warehouse, SQL marts and…
A collaborative-filtering movie recommender using from-scratch SGD matrix factorization on real MovieLens 1M data, served via FastAPI.
An e-commerce recommendation engine combining a from-scratch ALS implementation (NumPy/SciPy) with content-based scikit-learn filtering.
An NLP triage classifier for political statements trained on the real LIAR fake-news dataset, with literature-comparable, interpretable results.
A support-ticket triage platform with a Gradio UI and MLflow experiment tracking for model iteration.
Automated visual quality control for manufacturing, using OpenCV/scikit-image feature extraction with scikit-learn classification.
Hourly bike-share rental demand forecasting with honest time-series validation, SHAP explainability, and a containerized FastAPI serving layer.
Analytics engineering on a real 99,441-order e-commerce dataset: a dbt star schema on DuckDB, 34 automated data tests, and a…
A dbt + DuckDB ELT pipeline over synthetic retail data, with 35 automated data-quality tests and a Streamlit dashboard.