Joshua Ebinisi

Selected as one of 50 participants for JPMorganChase's 24-hour Data for Good hackathon. Our team worked with the National Basketball Players Association (NBPA) on a real question: how do travel and scheduling affect NBA teams on the court? The answer: it is not the miles, it is the missing rest.

• Rebuilt every trip for 6,150 regular-season games across 51 arenas: miles, rest days, time zones, altitude, and road-trip length.
• Same team, one night apart: teams win 51% on night 1 of a back-to-back and only 44% on night 2.
• A tired team facing a rested one won just 38% of the time, and a tired home team's win rate fell from 63% to 47%.
• Led the design of the NBPA-styled dashboard, including a schedule heat map that scores every team's 2026-27 week from Calm to Brutal.

I worked on PNC's Data Modeling and Analytics (DMA) team, where the group was responsible for the bank's inventory of quantitative models. Work spans regulatory forecasting, risk intelligence automation, and large-scale data engineering.

• 1,000 Monte Carlo simulations behind a CCAR/DFAST loss-forecasting model, testing a 90% prediction interval against actual quarterly results, deployed with Python's pickle.
• Productionized a regulatory and risk intelligence pipeline in Python that surfaces daily reporting for Model Risk, MRMG, Audit, and OP Risk, driving +10% development output.
• Engineered an end-to-end Hadoop pipeline (PySpark · Hive · Impala) flagging client spending spikes to trigger banker outreach, scheduled via Slurm/PowerShell.