Principal Machine Learning Engineer (EO/IR Image Analysis)

Description

The company builds rockets, spacecraft and critical subsystems, including solar panels, flight software, star trackers, optical systems, separation systems and radios.

Its Space Systems division has supported more than 1,700 missions, from interplanetary exploration and in-space manufacturing to national security and defense work. Hardware from the division has flown to the Moon and Mars, and its customers include NASA, the Space Development Agency and the U.S. Space Force.

The Optical Systems group provides sensing solutions to Department of Defense and Intelligence Community customers for space domain and ISR problems across orbital, terrestrial and airborne settings. It draws on more than 20 years of electro-optical and infrared systems work.

The role is based in Tucson, AZ and supports Tranche 3 of the Space Development Agency's Proliferated Warfighter Space Architecture and later work, by building deep learning neural networks for EO/IR image processing.

Responsibilities

• Build deep neural networks and computer vision algorithms for image and video processing pipelines.
• Design, train and deploy machine learning models for optical systems.
• Use physics-based image simulation frameworks to scale training and validation data.
• Rapidly prototype and further develop solutions that will be integrated into operational systems.
• Troubleshoot, fix bugs in and maintain existing and new systems.
• Work with other researchers and engineers on AI, machine learning and computer vision.
• Track ML research and assess new tools, frameworks and techniques.

Requirements

• A bachelor's degree and 12+ years of experience, or a master's degree or PhD in computer science, electrical and computer engineering, mechanical engineering, physics or a related field.
• Strong grounding in ML fundamentals, including model selection, architecture, training, validation, testing and deployment.
• Experience building CNN-style deep learning networks for computer vision, image processing or analysis, such as object detection, image segmentation and tracking.
• Strong math background, particularly linear algebra and optimization theory.
• Proficiency in Python and in deep learning libraries such as Keras, TensorFlow and PyTorch.
• Software engineering fundamentals: version control, testing, CI/CD and containerization.
• Proficiency in C, C++ and Rust.
• Experience curating high-quality, real-world datasets for training deep learning models.
• TS/SCI security clearance.
• U.S. citizenship is required due to program requirements.
• Ability to obtain and maintain a U.S. Government security clearance.

Conditions

Based in Tucson, AZ.

Works to tight deadlines and under pressure, with the focus on delivering results.

May require prolonged periods of sitting, standing, walking and computer work, with occasional exposure to moderate noise, dust and fumes in production areas.

For US offices: under ITAR export rules, you must be a U.S. citizen, a lawful U.S. permanent resident, a refugee or asylee, or eligible to obtain the required authorizations from the U.S. Department of State and/or Department of Commerce.

For New Zealand offices: background checks, including nationality checks, are run before any offer is made. You must be eligible to access ITAR-regulated equipment and data. You may be ineligible if you do not hold citizenship of Australia, Japan, New Zealand, Switzerland, the European Union or a NATO country, or if you hold ineligible dual citizenship or nationality.