• Led project migrating legacy reports to Power BI for Canadian government housing agency, dealing with very large and sensitive data• Led project migrating client’s legacy reports from Sisense Insights to Power BI, including full migration of KPIs as well as improving data model performance with DAX• Developed and maintained ETL pipeline from company data, for Power BI reports on consultants’ performance to management, improving billable efficiency by 15%• Improved QA… Show more • Led project migrating legacy reports to Power BI for Canadian government housing agency, dealing with very large and sensitive data• Led project migrating client’s legacy reports from Sisense Insights to Power BI, including full migration of KPIs as well as improving data model performance with DAX• Developed and maintained ETL pipeline from company data, for Power BI reports on consultants’ performance to management, improving billable efficiency by 15%• Improved QA of invoice billing runs with Power BI dashboards to identify inconsistent item charges; increased rate of finding errors by 10%• Managed Power BI workspaces, licensing, data refresh, and embedded reports into third-party CRMs and ERPs• Created large-scale, near real-time Power BI reports; from production-line KPIs, to invoice-generation QA with millions of records per billing cycle • Automated custom CRM reports with Python/Go and SQL logic to help clients gain insights• Integrated Power BI with Dynamics CRM, helping sales team track opportunity pipelines and sales funnels• Processed, cleaned, and verified integrity of data, and engineered features for predictive models• Trained and maintained predictive models for company’s billable efficiency to provide management with forecasted bottom line and identify associates with need for improvement• Conducted regular client meetings and supported successful implementation of data-driven solutions Show less

• Led ROP (Research Opportunity Program) undergraduate team in researching effects of medical imaging data heterogeneity on CNN (convolutional neural network) performance• Created, configured, and managed virtual environments (Anaconda) on Ubuntu server • Managed data of experiment results in PostgreSQL database (backups and updates)

Department of Medical Imaging• Led team of 6 authors for PubMed research paper 'Quantifying Uncertainty in Machine Learning Classifiers for Medical Imaging' • Modelled predictive uncertainty metric for transfer-learned ConvNet with PyTorch using NumPy• Reduced classification uncertainty at test-time by 17% by balancing dataset• Cleaned and pre-processed imaging data for team of 5 researchers for PubMed research paper 'Assessing and Mitigating the Effects of Class… Show more Department of Medical Imaging• Led team of 6 authors for PubMed research paper 'Quantifying Uncertainty in Machine Learning Classifiers for Medical Imaging' • Modelled predictive uncertainty metric for transfer-learned ConvNet with PyTorch using NumPy• Reduced classification uncertainty at test-time by 17% by balancing dataset• Cleaned and pre-processed imaging data for team of 5 researchers for PubMed research paper 'Assessing and Mitigating the Effects of Class Imbalance in Machine Learning with Application to X-ray Imaging' • Used NumPy and Pandas for data pre-processing Show less

• Constructed machine learning pipeline (pre-processing, feature extraction, ConvNet) to classify joint bleeds in ultrasound imaging data using PyTorch• Maintained project workflow with Git, GitHub, and Jupyter Notebook• Implemented Naive Bayes algorithm in Scikit-Learn to parse .csv file and label radiologist text in a semi-supervised way to augment dataset

• Collaborated in the department of medical imaging on ML pipeline to help dental radiologists replace worn-out dental PSP plates based on wear-and-tear• Pre-processed image data by creating a Python program to enhance image contrast • Standardized image data prior to feature extraction via PCA in Scikit-Learn• Used DBSCAN to segment out X-ray image artefacts and hand-engineer features to cluster images by K-means algorithm

Teaching assistant for MAT133: Calculus and Linear Algebra for Commerce• Developed public speaking skills by teaching two weekly tutorial sections of 35 students each• Helped students improve their test and exam scores by at least 20% through my instruction in both classroom and private office hour settings