Full Stack Developer (Python, Java, Docker, Kubernetes)

As a developer for the US Census Bureau, you will design, develop, and maintain the highly complex routing and scheduling algorithms that power the Mojo field operations control system. You will architect modular, scalable optimization services capable of handling both massive batch schedules and real-time, on-demand routing. You will ensure maximum operational efficiency and data-driven accuracy for critical national survey workflows.

Optimizer Design, Development & Support • Design and implement vehicle routing and scheduling algorithms using OR-Tools to generate optimal field assignments • Develop clustering algorithms to group workloads geographically and minimize travel time/cost • Implement constraint models (capacity, time windows, skills, priorities, survey rules) within the optimizer • Integrate the optimizer with the Mojo control system APIs for job intake, execution, and results publishing • Build batch and real-time optimization modes to support both scheduled and on-demand routing • Refactor optimizer components into modular, extensible strategy interfaces for new routing heuristics • Implement data preprocessing pipelines to normalize locations, distances, and travel matrices • Develop automated test harnesses validating solution correctness against known datasets and constraints • Implement “solution quality” scoring metrics (distance, balance, SLA adherence, cost) to evaluate goodness of routes • Create regression benchmarks comparing new algorithm performance vs. baseline outputs • Add parallel processing and scaling support to handle large route sets and high agent counts • Instrument services with logging and metrics to track runtime, solver performance, and solution quality • Package optimizer services for CI/CD deployment with reproducible builds and environment configs • Document routing logic, constraints, integration points, and operational runbooks • Provide production support, tuning, and continuous improvement of optimization heuristics and performance