Job Description

Swarm Engineer - Multi-Agent Task Planning

• Location: Phoenix · on-site
• Company Stage: Innovative startup in swarm robotics
• Office Type: on-site
• Salary: $150K–$160K

We're representing a pioneering company in the swarm robotics space, focusing on transforming the defense industry with low-cost, autonomous robots. Their cutting-edge approach combines advanced machine learning with multi-agent coordination, aimed at creating swarms of Unmanned Ground Vehicles (UGVs) that can execute complex tasks in real-time, enhancing operational capabilities and saving lives.

What You Will Do

• Architect, train, and iterate on multi-modal action models to enable UGVs to select and execute swarm-level tactical macro-actions from diverse inputs.
• Design model architectures that integrate local perception, swarm state, and mission objectives into a cohesive decision-making framework.
• Develop and implement reinforcement learning techniques, including transformer-based models, to optimize swarm coordination policies.
• Optimize machine learning models for real-time deployment on edge devices through advanced techniques like quantization and distillation.
• Build and maintain the complete pipeline, from data collection and training to evaluation and field deployment of models.
• Collaborate with cross-functional teams to integrate action models into the overall autonomy stack, ensuring seamless operation in field environments.

Ideal Candidate Background

• Strong foundation in neural networks, reinforcement learning, and multi-agent decision-making frameworks.
• Proficient in Python and C++, with hands-on experience in PyTorch or TensorFlow.
• Demonstrated experience in training models that produce actions (e.g., reinforcement learning, planning-as-inference).
• Familiarity with multi-agent coordination concepts, including task allocation and swarm behaviors.
• Experience optimizing ML models for deployment on resource-constrained or edge hardware.

Preferred

• Hands-on experience with policy gradient methods such as PPO.
• Knowledge of multi-agent task planning algorithms and tools like ONNX or TensorRT.
• Background in robotics, autonomous systems, or simulation environments.

Compensation and Benefits

This position offers a competitive base salary of $150K–$160K, along with exciting opportunities for growth and innovation in a dynamic startup environment.