Applied Robotics Engineer: Real-World AI Deployment

Medal.TV; Washington (District of Columbia); On-site; USD 150,000 - 210,000; Full time

Job summary

Medal.TV is seeking an Applied Robotics Engineer to bridge research with deploying AI systems in the real world. You will deploy models on hardware under constraints like limited compute and noisy sensors, focusing on post-training evaluation, system integration, and field deployment.

You will partner with customers to understand legacy control systems, sensor pipelines, and edge compute limits, and you will work across the hardware-software boundary in real-time environments.

Qualifications

• Deploy AI or robotics systems from prototype to production.
• Own models end-to-end from training to real hardware, including failures.
• Debug and fix root causes rather than symptoms in complex systems.
• Strong Python and PyTorch; C++ when needed for performance.
• Experience with quantization, distillation, TensorRT or ONNX export.
• Deep knowledge in RL/control, world models, or perception.
• Experience across sensors, middleware, edge compute, and real-time constraints.
• Rigorous evaluation; understands what benchmarks can and cannot tell.

Responsibilities

• Bridge research with real-world hardware deployment and evaluation.
• Work with partners to understand legacy systems, sensor pipelines, latency, and edge compute limits.
• Handle post-training evaluation, system integration, and deployment on hardware.
• Diagnose and fix failures in hardware-in-the-loop deployments rather than chasing symptoms.
• Collaborate with engineers and leadership in customer settings to explain model capabilities.

Skills

Python & PyTorch

C++ for performance

Model deployment

Quantization & TensorRT

RL / Immitation learning

Perception & sensor fusion

Robotics middleware

Real-time systems

Benchmarking & evaluation

Job description

Medal.TV is seeking an Applied Robotics Engineer to bridge research with deploying AI systems in the real world. You will deploy models on hardware under constraints like limited compute and noisy sensors, focusing on post-training evaluation, system integration, and field deployment.

You will partner with customers to understand legacy control systems, sensor pipelines, and edge compute limits, and you will work across the hardware-software boundary in real-time environments.