Machine Learning Engineer

We're seeking a skilled Machine Learning Engineer to build and deploy production ML systems for the next-generation data management and artificial intelligence platform for maritime domain awareness.

Spear AI is a growing defense contracting company dedicated to delivering cutting-edge solutions that support our nation's security. As we expand, we're building a culture where innovation meets mission-critical work. We operate with a flat organizational structure that empowers every team member to make an impact, collaborate directly with leadership, and contribute to projects that matter. Whether you're joining our Hardware, Software, or Services division, you'll work alongside talented professionals who are committed to excellence and advancing the capabilities that keep our nation safe and secure.

Spear AI builds sonobuoy sensors that are deployed into the water and collect edge data. We also work with the U.S. Navy to collect and process their SONAR data. You'll have an opportunity to work on real-world projects that directly impact warfighter capabilities and mission success.

What You'll Do

• We're a small team wearing many hats, and you'd have a wide variety of responsibilities that include:
• Design, train, and optimize machine learning models using PyTorch
• Deploy models to production environments in the cloud and at the edge
• Build and maintain ML pipelines for training, evaluation, and inference
• Integrate machine learning models into real-time and batch processing systems
• Optimize model performance for accuracy, latency, and resource constraints
• Implement model monitoring, versioning, and deployment strategies
• Work with signal processing data and time-series analysis
• Improve local development and CI/CD for ML workflows using modern tooling and GitHub Actions

Who You Are

• We're looking for someone with strong Machine Learning Engineering skills who shares our most important values:
• You're fanatical about polish. Every detail matters. You love to make sure your code is linted, formatted, fully typed, and has comprehensive test coverage.
• You care about correctness. You take pride in the fact that your models perform reliably and downstream consumers trust your predictions.
• You obsess over performance. You daydream about model latency, throughput, and efficient inference pipelines.
• You dive deep. It's important for you to really know how things work. You're always building prototypes and setting up experiments to reinforce your understanding.
• You live on the bleeding edge. You've got a long list of upcoming ML techniques and frameworks you're excited about and can't wait to experiment with new approaches.