Ariel Boyd
Data Analyst | Finance, E-Commerce, Business Analytics
New to Contra
Ariel is ready for their next project!
An interactive D3.js dashboard analyzing revenue, churn, and customer segments for a telecom subscription business (7,043 real customers). Calculates "Revenue at Risk" turning a churn rate into a churn cost and finds that contract type is the strongest churn predictor (42.7% for month-to-month vs. 2.8% for two-year contracts), while payment method is a surprisingly strong signal on its own: electronic check users churn at 45.3%, nearly 3x automatic payment methods.
2
24
An interactive D3.js dashboard exploring 502 S&P 500 companies across 11 sectors, with 10 years of index price history. Four linked, coordinated views (line chart, sector breakdown, market cap ranking, and a growth-vs-price scatter plot) let users cross-filter by sector and drill into individual companies. Key findings: Technology dominates the index with 82 companies, nearly 4x Energy or Basic Materials, and just three companies (AAPL, NVDA, MSFT) represent roughly $10.3T in market cap, more than several entire sectors combined.
2
43
A Python data pipeline analyzing how BMW, Mercedes-Benz, and Audi depreciate compared to Ford, using 87,000 real Craigslist listings. Caught and fixed a major data quality issue (55% of raw data was duplicate listings), quantified value retention by brand, and trained a Random Forest model predicting price within $4,200 average error. Audi retains value best of the three luxury brands at every time horizon.
2
59
A Python data pipeline analyzing oil, gold, and silver price trends, volatility, and correlation using the yfinance API. Found gold and silver move together strongly (0.78 correlation) while oil is independent, and that silver carries nearly 2x gold's volatility despite both being "precious metals." Built to inform my own trading research.
0
31