Spécialiste, Machine Learning (SSA) - Machine Learning Specialist (SSA)

Northstar Earth and Space•Montreal, QC

136d•Hybrid

About The Position

NorthStar Earth & Space (“NorthStar”) is the world’s leading space-based data analytics company. We use optical, radar, passive RF and other Space Situational Awareness (SSA) data as building blocks to create information services tailored to the needs and requirements of clients from both government and industry. Our extensive and patented technology portfolio includes orbit dynamics and machine learning algorithms for a variety of applications for space domain awareness and space traffic management. With headquarters in Montreal, Canada, European headquarters in Luxembourg, and US operations in McLean, Virginia, the company is solving the ever-increasing threat of space collisions and empowering humanity to preserve our planet. The Space Information and Intelligence (Si²) team is a multidisciplinary team of scientists, engineers and software developers with backgrounds in Physics, Engineering Dynamics, Computer Science and Software Design. The team is dedicated to developing innovative solutions for monitoring the space environment using multiple sources for observations and geospatial data.

Requirements

MSc or PhD in Machine Learning, Physics, Electrical/Computer Engineering, Applied Mathematics, Aerospace Engineering, or related fields.
5+ years of experience
Bilingualism French & English spoken and written (collaboration with international colleagues - 20%)
Any combination of education and relevant experience may be considered
Strong foundations in machine learning, probability, statistics, and optimization.
Hands-on experience with deep learning frameworks (PyTorch for example) and Python in general and its scientific libraries with experience implementing and debugging custom model components.
Strong deep learning architecture knowledge with understanding of convolutional and attention-based architectures, with practical experience adapting pretrained models or building them from first principles.
Ability to design proof-of-concept with production constraints in mind (modularity, scalability, reproducibility) with solid software engineering practices (Git, testing frameworks, CI/CD).

Nice To Haves

Experience working in a multidisciplinary team deploying models within data pipelines whether edge and/or cloud environments.
Knowledge of high-performance computing (CUDA, GPU optimization).
Interest in or exposure to orbital mechanics.
Experience in large-scale data ingestion, preprocessing, and lifecycle management, including efficient handling of high-volume, high-dimensional datasets.

Responsibilities

Design, implement, and validate ML/DL models (supervised or unsupervised) to extract near-Earth astro-dynamical data from noisy imagery under limited hardware resource availability, identify time series representing the dynamic of an object, cluster time series, infer astro-dynamical data from partial observations, and analyze long-term trends and object behavior in orbit.
Support deployment of those models within a cloud infrastructure or on edge-devices.
Build robust, reusable software components beyond research prototypes and collaborate with a multidisciplinary team.
Benchmark solutions using simulated and real data.
Document algorithms, workflows, and results clearly for reproducibility and maintainability.