As Lead Analyst on a Virginia transportation safety study, I worked on a large-scale analytics pipeline that used connected-vehicle data to support more proactive safety evaluation than traditional crash-only analysis. The study covered 15,615,827,532 events across 188 days and integrated trajectory data with crash, roadway, geospatial, and census-derived context.

Key contributions

• Built data cleaning, processing, and geospatial analysis workflows for speeding, harsh braking, and harsh acceleration events
• Combined connected-vehicle data with roadway attributes, posted speed limits, crash records, bike-lane GIS layers, OpenStreetMap features, and demographic activity proxies
• Used spatial indexing and geospatial overlays to analyze both road segments and intersection contexts in Arlington and Fairfax Counties, Virginia

Impact

The project produced decision-relevant findings for infrastructure prioritization, targeted enforcement, and routine data-driven safety assessment. It also demonstrated my experience working with large-scale real-world data systems where spatial reasoning, data quality, and practical stakeholder interpretation all matter.