Senior AI Engineer (Closed-loop Simulation & RL)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI

Last checked: 1 hour agoStatus: Live

Job Summary

Build and extend a closed-loop evaluation framework for driving VLA/world model/planning systems, using log replay, scenario-based simulation, and regression testing to surface failure modes. Improve driving policies through reinforcement learning (including offline RL), reward modeling, policy distillation, and post-training. Define safety/comfort/progress metrics, run simulation rollouts and failure analysis pipelines, and partner with VLM/Data/ML Platform teams to integrate closed-loop training and evaluation.

Location: South Korea

Workplace: Hybrid

Employment Type: Full time

Job Function: Data Science & Machine Learning

Seniority: Mid level

Key Responsibilities

•Design and develop a closed-loop evaluation framework for driving VLA/world model/planning systems.
•Build pipelines for log replay, scenario-based simulation, regression tests, and safety-critical case mining.
•Improve driving policies using RL/offline RL, reward modeling, policy distillation, and post-training.
•Design and operate simulation rollout, reward functions, evaluation metrics, and failure analysis pipelines.
•Define driving quality indicators (safety, comfort, progress, rule compliance) and develop improvement loops using counterfactual simulation, synthetic scenarios, and real-world driving logs, in collaboration with VLM/Data/ML Platform teams.

Key Requirements

•5+ years of research/development experience (or equivalent) in reinforcement learning, imitation learning, planning, simulation, robotics, or autonomous driving.
•Hands-on experience with at least one of closed-loop evaluation, simulation rollout, scenario testing, or policy evaluation.
•Experience developing ML training or evaluation pipelines using Python and PyTorch.
•Understanding of RL, reward modeling, trajectory planning, and sequential decision making.
•Proven ability to analyze failure cases and improve metrics, rewards, datasets, and training recipes.

Experience:5+ yearsAutonomous drivingRobotics

Skills:CollaborationProblem-solvingData analysis

Tech Stack:PythonPyTorchReinforcement learningOffline RLReward modelingPolicy distillationPost-trainingTrajectory planningVLMVLAWorld ModelPlanningLog replaySimulationScenario-based simulationRegression testingCounterfactual simulationSynthetic scenarioWaymaxCARLA

Company Brief

42dot

Website

Loading job details...

Preparing the role view and application actions.

42dot

1 week ago

Senior AI Engineer (Closed-loop Simulation & RL)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI

Last checked: 1 hour agoStatus: Live

Job Summary

Location: South Korea

Workplace: Hybrid

Employment Type: Full time

Job Function: Data Science & Machine Learning

Seniority: Mid level

Key Responsibilities

•Design and develop a closed-loop evaluation framework for driving VLA/world model/planning systems.
•Build pipelines for log replay, scenario-based simulation, regression tests, and safety-critical case mining.
•Improve driving policies using RL/offline RL, reward modeling, policy distillation, and post-training.
•Design and operate simulation rollout, reward functions, evaluation metrics, and failure analysis pipelines.
•Define driving quality indicators (safety, comfort, progress, rule compliance) and develop improvement loops using counterfactual simulation, synthetic scenarios, and real-world driving logs, in collaboration with VLM/Data/ML Platform teams.

Key Requirements

•5+ years of research/development experience (or equivalent) in reinforcement learning, imitation learning, planning, simulation, robotics, or autonomous driving.
•Hands-on experience with at least one of closed-loop evaluation, simulation rollout, scenario testing, or policy evaluation.
•Experience developing ML training or evaluation pipelines using Python and PyTorch.
•Understanding of RL, reward modeling, trajectory planning, and sequential decision making.
•Proven ability to analyze failure cases and improve metrics, rewards, datasets, and training recipes.

Experience:5+ yearsAutonomous drivingRobotics

Skills:CollaborationProblem-solvingData analysis

Company Brief

42dot

Website