Software Engineer Intern
Cambridge Mobile Telematics · June 2026 – August 2026 · Cambridge, MA
- Improved performance for the telematics analytics API by caching repeated table initialization for 6 hours, ∼35% fewer database calls and faster dashboard and chart queries across large customer datasets.
- Built and deployed MCP tools with Terraform that let an AWS DevOps agent query a time-series metrics database for anomalies, reducing root-cause analysis time by ∼80% for ride failures and device-update issues.
- Built a lightweight Python polling system for an analytics chart server that refreshed data every minute, served 1,000+ daily chart requests, and reduced repeated backend requests substantially reducing computational costs.
- Built a simple rule based anomaly-detection algorithm in Python using pandas and DuckDB to scan 400,000+ driver records across rolling 3-day windows and surface ride-failure patterns for further investigation.
