Intermediate Python programming, APIs, exploratory data analysis, statistical modeling, machine learning foundations, and collaborative software development.
About
DS 2500 focused on intermediate Python programming for data science — object-oriented design, working with APIs, cleaning messy real-world datasets, building visualizations, and introducing machine learning fundamentals within a collaborative software engineering workflow.
The semester culminated in a team research project analyzing consumer banking complaints using live CFPB data. We collected roughly 35,000 records through the CFPB API, explored complaint patterns across major institutions, and evaluated whether regulatory enforcement actions appear to change harmful consumer outcomes — my first complete end-to-end data science pipeline.
Skills
Featured Project
Do Regulatory Penalties Change Bank Behavior?
Using roughly 35,000 CFPB consumer complaints collected through the CFPB API, our team investigated whether major enforcement actions against large banks appear to reduce harmful consumer complaint patterns. The project combined API-based data collection, exploratory analysis, visualization, and statistical comparisons across multiple financial institutions.
Project Preview
Downloads
Full written research report on regulatory penalties and consumer complaint patterns across major banks.
Open PDFTeam presentation summarizing methodology, findings, and conclusions for the capstone project.
Open PDFOriginal project requirements, deliverable expectations, and grading criteria for the semester capstone.
Open PDFDS 2500 syllabus covering Python programming, data analysis, machine learning, and project milestones.
Open PDFLearning Outcomes
Querying external data sources programmatically and handling paginated JSON responses.
Parsing, filtering, and restructuring nested API payloads into analysis-ready tables.
Communicating patterns clearly with Matplotlib charts tuned for technical audiences.
Investigating distributions, outliers, and relationships before formal modeling.
Designing reusable Python classes that organize data pipelines and project logic.
Writing intermediate-level scripts with functions, modules, and disciplined structure.
Applying foundational ML ideas to real datasets with appropriate skepticism and validation.
Dividing responsibilities, integrating contributions, and shipping a shared codebase.
Framing questions, citing sources, and defending conclusions with evidence.
Branching, merging, and reviewing teammate changes on a shared repository.
Comparing groups, interpreting variation, and avoiding overclaiming from noisy data.
Distilling complex analysis into slides that non-specialists can follow.
Reflection
DS 2500 significantly strengthened my programming abilities beyond introductory Python. I learned to work with real-world APIs, clean messy datasets that never arrive in tidy CSV form, and build reproducible analysis pipelines my teammates could trust. Creating professional visualizations and communicating technical findings through written reports and presentations became as important as writing the code itself.
The CFPB banking project was my first complete end-to-end data science workflow — from API collection through exploratory analysis to a defended conclusion. Collaborating with Git, dividing analytical responsibilities, and integrating our work into a single narrative taught me how data science actually happens in teams. That foundation directly prepared me for future machine learning and analytics coursework, where rigor, reproducibility, and clear communication are non-negotiable.