Application close date:
Applications will be accepted on an ongoing basis until the requisition is closed.
At Blue Origin, we envision millions of people living and working in space for the benefit of Earth. We’re working to develop reusable, safe, and low-cost space vehicles and systems within a culture of safety, collaboration, and inclusion. Join our team of problem solvers as we add new chapters to the history of spaceflight!
This role is part of Blue National Security (BNS), where our focus is the design, development, and delivery of launch and spacecraft systems for national security missions. Through Blue Ring, our all-in-one, high-powered hybrid solar-electric and chemical-propelled spacecraft, we provide unmatched mission maneuverability, versatility, and capacity at dramatically lower-profile costs.
Blue National Security | Autonomy & Intelligent Mission Products
Within the Autonomy & Intelligent Mission Products organization, this team operates at the forefront of AI-enabled products, developing next-generation products that address the most demanding and consequential challenges in the defense domain.
Special Mentions
Relocation assistance is available for this position
Travel expectation: up to 10%
This is an individual contributor role
Ability to obtain a TS/SCI clearance
Job Description
Why this role Turning radar returns into a confident answer about what an object is and how it behaves is still an open problem. On this team, you'll own the AI layer that takes raw radar data, finds objects of interest, and characterizes them well enough that the autonomy stack or operator can act on the result. You will be developing technology where labeled data is scarce, hardware resources are constrained, and the output matters. This is a new development program where you get to decide the model architecture.
Responsibilities include but are not limited to:
Design, train and evaluate deep learning models (GNNs, transformers, DRL, Learned Algorithm Unfolding or hybrids) that find and classify objects of interest in raw radar data as well as range-Doppler and range-angle maps.
Build characterization models that estimate object attributes such as size, shape, motion state, track, behavior, with calibrated uncertainty.
Represent radar data in the right form for each model, for example point or track graphs for GNNs.
Build the data engine: synthetic data generation, simulation-based training, labeling strategy, augmentation, and handling of class imbalance.
Detect inputs that fall outside the training data, and flag low-confidence results so downstream systems know when not to trust the model.
Work with the Advanced Systems, DSP, GNC, and flight software teams to define how data is handed between stages and to deploy models on embedded or edge hardware.
Define the metrics, test sets, and evidence needed to qualify ML models for deployment to the edge.
Minimum Qualifications
BS in Computer Science, Electrical Engineering, Computer Engineering, Physics, or a related field, plus about 3 years of ML engineering experience.
A track record of building and shipping deep learning models in PyTorch (or similar).
Edge deployment (TensorRT, quantization, FPGA or SoC targets).
Strong Python skills
Working knowledge of radar fundamentals: range, Doppler, angle, resolution, clutter, and multipath.
Preferred Qualifications
Familiarity with Ground Moving Target Indicator (GMTI) or Airborne Moving Target Indicator (AMTI) signal processing concepts and their implications on system design
Experience with sensor data from radar, sonar, SAR, lidar, or RF.
Training on synthetic or simulated data, and closing the gap between simulation and real data.
Uncertainty quantification, calibration, or out-of-distribution detection.
Experience with embedded C++
Active TS/SCI
What you get in this role
Ownership from the start. You'll define the radar AI architecture on a new program.
Work that matters. Your models will directly support national security…