Data Analyst · Data Scientist — SQL, Python & Power BI
I build end-to-end data solutions — from SQL-based analysis and ETL pipelines to KPI dashboards and machine learning models — that turn raw, messy data into decisions people can act on.
About
I'm a Data Analytics and AI professional based in the Greater Toronto Area, with a PG Diploma in Artificial Intelligence with Machine Learning from Humber Polytechnic and a PG Diploma in Cloud Operations from York University.
My work has focused on campaign measurement analytics, ROI reporting frameworks, and root-cause analysis that feed directly into process and automation improvements.
I'm currently seeking Data Analyst, Data Scientist, or ML Engineer roles across Canada — available to start immediately.
Skills
Projects
An end-to-end analytics build on MySQL, Python and Power BI, processing 100K+ support tickets through 30+ SQL queries for KPI monitoring and anomaly detection.
A YOLOv5/YOLOv9-based computer vision system built as my Humber capstone for real industry client Kevares Inc. — detects and blurs faces and license plates from autonomous robot camera feeds. Best model (YOLOv5l): 90.6% precision, 86.5% recall, 87.5% mAP50.
A multi-modal deep learning model trained on 5,000+ image-caption pairs, reaching 85%+ validation accuracy. Reduced training loss by 25% through hyperparameter tuning.
An Excel analytics dashboard built on 10,345 job postings, with 9 charts and KPI cards across 4 sheets.
Built a transit service reliability analysis on 23,701 real TTC delay records, engineering a station-name normalization function with a 99.6% match rate and comparing Random Forest vs. Gradient Boosting models (~0.62–0.65 ROC-AUC) across a 5-feature experiment set, including a leakage-safe rolling delay frequency feature.
Built a causal inference case study measuring the true incremental impact of a content intervention on engagement, using an X-learner uplift model with Qini AUC scoring (0.146 ± 0.034 cross-validation) and a permutation placebo test, presented in a 7-tab interactive dashboard.
Completed a virtual Data Scientist job simulation at BCG X, working on a customer churn case for a simulated energy utility client (PowerCo). Conducted EDA on ~14,600 customers, engineered price-sensitivity and consumption-based features, built a Random Forest classifier to predict churn, and delivered an executive summary to simulated senior stakeholders.
Experience
Education
Certifications
Contact
Available to start immediately, open to opportunities across Canada.