Transportation Optimization Algorithm Engineer

We are hiring a Transportation Optimization Algorithm Engineer to own, evolve, and extend the routing engine at the heart of Sophus's platform. Our current algorithm is built on JSprit — a Java-based Vehicle Routing Problem (VRP) solver — and uses Large Neighborhood Search (LNS) as its core optimization mechanism. What You Will Do Algorithm Development & Extension • Design, implement, and test extensions to the existing JSprit-based routing engine, ensuring each change is evaluated against the full algorithm behavior • Translate incoming customer requirements (new constraints, new objectives, new problem types) into precise algorithmic changes, mapping out second-order effects before writing a single line of code. • Develop and tune LNS operators (destroy and repair heuristics), including ruin-and-recreate strategies, regret-based insertion, and adaptive operator selection. • Extend or replace existing heuristics where necessary, considering exact methods (e.g., branch-and-bound, column generation) as sub-routines for structured sub-problems. What We Are Looking For Required • Language: JAVA • JSprit: Hands-on experience with JSprit or deep familiarity with its internals (problem definition, solution representation, acceptors, selectors, and custom algorithms) • LNS / Metaheuristics: Practical experience implementing or tuning Large Neighborhood Search (LNS) — specifically destroy/repair operators, adaptive weight adjustment, and acceptance criteria • VRP Fundamentals: Solid understanding of the Vehicle Routing Problem (VRP) family: VRP, CVRP, VRPTW, PDPTW, heterogeneous fleets, multi-depot, open routes • Architecture Mindset: Systems-level thinking: ability to reason about how any single algorithmic change propagates through constraint interactions, solution quality, and runtime behavior Strongly Preferred • Familiarity with other metaheuristics: simulated annealing, tabu search, genetic/evolutionary algorithms, ant colony optimization • Exposure to exact methods: branch-and-bound, branch-and-price, column generation — even as sub-routines within a heuristic framework • Academic or applied background in combinatorial optimization, operations research, or computer science with a focus on scheduling/routing • Experience with multi-objective optimization and Pareto-efficient solution sets Nice to Have • Familiarity with supply chain concepts: last-mile delivery, hub-and-spoke networks, time windows, load constraints • Experience in a SaaS product environment where algorithm changes ship to live customers • Python scripting for experiment analysis, visualization, or benchmark automation