Lodestar
Software Engineer I: Perception
About this role
Join Lodestar's perception team as a Software Engineer I to develop computer vision and machine learning algorithms that detect and classify space targets in real time for autonomous spacecraft defense systems. You'll work across the full lifecycle of perception models—from research through deployment—collaborating with senior engineers in a fast-paced, early-career growth environment.
What you'll do
- Contribute to perception algorithms at the core of MITHRIL's autonomy software suite
- Work through full ML lifecycle including literature review, training, evaluation, optimization, and deployment
- Build and optimize models for edge computing and develop new perception components
- Develop methods for predicting relative pose of dynamic space targets in challenging visual environments
- Integrate perception models into mission simulation and real-time autonomy pipelines
- Maintain perception tooling, training infrastructure, and evaluation frameworks
What they're looking for
- Python and C++ software development
- Deep neural networks and machine learning fundamentals
- Computer vision and image processing (OpenCV, etc.)
- ML model training, evaluation, and optimization
- Edge/real-time ML deployment
- Linux, Git, and CI/CD pipelines
- Docker and containerization
- GPU acceleration frameworks (CUDA, TensorRT)
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Lodestar
Lodestar develops MITHRIL, an AI-powered autonomy suite for autonomous spacecraft operations that combines state estimation, perception, and real-time decision-making capabilities. The company is hiring Software Engineers specializing in state estimation, perception algorithms, and on-board autonomy to build advanced systems for space target detection, trajectory prediction, and autonomous mission execution.
View all jobs at LodestarLikely interview questions
- Walk us through an ML project where you took a model from training to evaluation—what challenges did you encounter?
- Describe your experience deploying models to resource-constrained or real-time systems. What optimization techniques did you use?