Sr. Reinforcement Learning & Autonomous Decision Systems Engineer

Aurex • Huntsville, AL

5d • Onsite

About The Position

Aurex is seeking a Senior Reinforcement Learning / AI Engineer to develop reinforcement-learning and AI-enabled decision systems for complex aerospace and defense applications. This role is centered on intelligent agents that make closed-loop decisions over time in simulation and, ultimately, in mission-relevant real-time environments. The work may include continuous control, discrete and hybrid decision spaces, planning, coordination, and decision-making under uncertainty and partial observability. The successful candidate will formulate decision problems, design learning environments, train and evaluate agents, and integrate learned policies with physics-based models and operational simulations. This is not primarily a perception or computer-vision role; the emphasis is on sequential decision-making, autonomous behavior, and rigorous engineering evaluation.

Requirements

• Bachelor’s degree in Computer Science, Computer Engineering, Aerospace Engineering, Electrical Engineering, Mechanical Engineering, Physics, Applied Mathematics, or a related technical field.
• Ten or more years of relevant professional experience in reinforcement learning, autonomy, machine learning, robotics, control systems, modeling and simulation, or related engineering disciplines. Additional relevant education may substitute for experience.
• Meaningful hands-on experience developing, training, and evaluating reinforcement-learning agents for sequential decision-making, planning, control, or autonomous-system applications.
• Strong Python software-development experience.
• Practical experience with at least one modern deep-learning framework, such as PyTorch, JAX, or TensorFlow.
• Experience creating or adapting simulation environments for learning agents, including defining observations, actions, objectives or rewards, constraints, scenarios, and evaluation metrics.
• Strong understanding of core reinforcement-learning concepts, including exploration, credit assignment, policy evaluation, training stability, generalization, and agent-environment interaction.
• Experience working with continuous, discrete, or hybrid decision problems.
• Experience with decision-making under uncertainty, stochastic environments, or partial observability.
• Experience integrating learned agents, algorithms, or software services with physics-based models, simulations, test harnesses, or larger software systems.
• Proficiency with modern software-development practices, including source control using Git, code reviews, automated or unit testing, software organization, and reproducible experimentation.
• Demonstrated ability to communicate complex AI, software, and engineering concepts to multidisciplinary technical teams.
• Ability to provide technical leadership and contribute effectively in a collaborative engineering environment.
• Active Secret security clearance or higher.
• Ability to work on-site at an Aurex office in Huntsville, Alabama.

Nice To Haves

• Master’s degree or Ph.D. in Computer Science, Aerospace Engineering, Electrical Engineering, Robotics, Applied Mathematics, Operations Research, or a closely related technical discipline.
• Advanced experience with modern reinforcement-learning methods, including actor-critic approaches, policy-gradient methods, value-based methods, offline RL, model-based RL, or hierarchical reinforcement learning.
• Experience with multi-agent reinforcement learning, cooperative or adversarial agents, distributed decision-making, or game-theoretic methods.
• Experience designing reinforcement-learning systems for aerospace, defense, autonomous vehicles, robotics, guidance and control, mission planning, battle management, or other safety- or mission-critical applications.
• Experience with distributed or large-scale RL training, including parallel simulation, distributed rollouts, GPU acceleration, cluster computing, or scalable experiment infrastructure.
• Experience with RL libraries or frameworks such as Ray/RLlib, Stable-Baselines3, CleanRL, TorchRL, Gymnasium, PettingZoo, or comparable internally developed frameworks.
• Experience integrating reinforcement learning with classical control, trajectory optimization, mathematical programming, search, planning, or model-predictive control.
• Knowledge of partially observable Markov decision processes, belief-state estimation, stochastic optimal control, or decision-making under uncertainty.
• Experience developing high-fidelity, physics-based, hardware-in-the-loop, software-in-the-loop, or distributed simulation environments.
• Experience with Monte Carlo analysis, design of experiments, uncertainty quantification, verification and validation, sensitivity analysis, or statistical performance assessment.
• Experience transitioning AI or autonomy algorithms from research or simulation environments into real-time or operational software systems.
• Familiarity with real-time software constraints, deterministic execution, latency management, fault handling, runtime assurance, or graceful fallback architectures.
• Experience with containerized and reproducible development environments using technologies such as Docker, Linux, CI/CD pipelines, or cloud/HPC computing environments.
• Experience leading technical efforts, mentoring engineers, defining technical approaches, or serving as a technical lead on multidisciplinary engineering programs.
• Experience supporting Department of Defense, intelligence community, aerospace, or other U.S. Government programs.
• Active Top Secret or TS/SCI security clearance.

Responsibilities

• Design, implement, train, and evaluate reinforcement-learning agents for mission planning, guidance and control, resource allocation, engagement management, battle management, and other autonomous decision problems.
• Translate operational and engineering problems into rigorous sequential-decision formulations, including states and observations; continuous, discrete, or hybrid action spaces; objectives and rewards; constraints; termination conditions; and uncertainty models.
• Build and maintain simulation-based learning environments that connect agents to vehicle, sensor, weapon, threat, environmental, command-and-control, guidance, navigation, and control models.
• Develop end-to-end training and evaluation workflows, including scenario generation, parallel rollouts, experiment tracking, checkpointing, regression baselines, reproducibility, and analysis of agent behavior.
• Train, tune, and debug agents, identifying issues such as training instability, poor exploration, reward misspecification, overfitting, weak generalization, and unintended exploitation of simulation behavior.
• Assess tradeoffs among model-free reinforcement learning, model-based learning, planning, classical control, optimization, and hybrid approaches, selecting methods based on mission and engineering requirements.
• Design evaluation campaigns to assess performance, robustness, generalization, uncertainty, edge cases, failure modes, interpretability, traceability, and operational relevance.
• Address real-time execution requirements, including inference latency, action constraints, deterministic interfaces, runtime monitoring, graceful fallback behavior, and integration with mission software.
• Use Monte Carlo analysis, sensitivity studies, trade studies, and controlled experiments to characterize agent performance and simulation assumptions.
• Collaborate with modeling and simulation engineers, software developers, systems engineers, analysts, and subject-matter experts to translate operational questions into executable learning and evaluation experiments.
• Apply modern software-engineering practices and AI-assisted development tools to accelerate prototyping, testing, refactoring, and documentation while maintaining engineering rigor.
• Provide technical leadership, mentor other engineers, and document architectures, methods, assumptions, interfaces, experiments, results, and recommendations.

Benefits

• Comprehensive total rewards approach to compensation
• Competitive benefits package designed to support employees' health, well-being, and professional growth.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior