CS/IDS 121: Relational Databases

Built a 96%-accuracy SVM for classifying Digital Experiences to Customer Journey Reasons. Improved Digital Steps classification from 46% to 100% for 94 million daily hits. Deployed pipelines with daily and weekly triggers in Azure DevOps and Azure Data Factory.

Used natural language processing (Bag-of-Words, Support Vector Machines) to classify lines of emails, achieving 94% recall and 96% precision. Modularized existing Python code into 8 modules, implementing version control and significantly reducing code size from over 1000 lines.