Home
Jobs
Constellation Space (YC W26)
Graduate/PhD Research Intern, Machine Learning

Graduate/PhD Research Intern, Machine Learning

On site

Seattle, United states

$ 7,000 /year

Junior

Internship

23-02-2026

Share this job:

Score my CV

Skills

Communication Python Rust Networking Research Training C++

Job Specifications

About Constellation

Constellation is building software to model, predict, and improve satellite network operations. We combine simulation, data/ML workflows, and product-facing platform systems to support better operational decisions.

Role Overview

We are seeking a research-focused ML intern (MS/PhD level) to help advance our modeling and experimentation capabilities. This role is ideal for someone who enjoys turning research ideas into rigorous experiments and high-quality prototypes that can influence real product and platform direction.

What You’ll Do

Design and run ML experiments for forecasting and anomaly/risk prediction in network operations
Develop and evaluate models using time-series, probabilistic, and simulation-informed approaches
Improve feature engineering, dataset quality, and evaluation methodology
Build reproducible research workflows for training, validation, and model comparison
Communicate findings through clear technical writeups and recommendations

What We’re Looking For

Currently enrolled in an MS or PhD program (CS, EE, Aerospace, Applied Math, or related)
Strong low level engineering skills and comfort with scientific/ML tooling (C++, Python, Rust)
Ability to own projects end-to-end: scoping, implementation, testing, and communication
Clear written/verbal communication and strong collaboration habits

Nice to Have

Experience with APIs, cloud infrastructure, or data-intensive systems
Familiarity with model evaluation, experiment tracking, and reproducibility
Background in networking, geospatial systems, telecom, or space-tech

About the Company

Constellation builds ConstellationOS, an AI-driven platform that makes satellite networks autonomous and resilient. We ingest telemetry from satellites, ground stations, and weather systems (100K+ messages/sec), predict link failures with 90%+ accuracy up to 5 minutes in advance, and automatically reroute traffic—zero data loss, zero human intervention. Know more