Reinforcement Learning Researcher

Key Skills

Agile methodologies, Algorithms, AWS, Azure, Cloud computing, Docker, Google Cloud Platform (GCP), Python Programming, PyTorch, Software Development, TensorFlow

Job Description

Reinforcement Learning Researcher

STR’s Analytics Division is seeking a reinforcement learning researcher to join our team of scientists and engineers, who develop and apply cutting edge reinforcement learning approaches for many problems, including swarm/drone deployment, fighter-jet dogfighting and tactical wargaming. Our projects integrate algorithmic research and software engineering to address challenging national security problems. We actively collaborate with research engineers across STR, as well as with academic researchers and industry partners to develop practical and powerful solutions to applying autonomy in the real world.

Our ideal candidate possesses knowledge of the state-of-the-art in reinforcement learning (RL), understands how to implement RL algorithms in software, and is only satisfied when their innovations become a reality by integrating them into real autonomous systems. Because we have a highly-collaborative, multi-disciplinary team, our perfect candidate also has a strong desire to learn about new, interesting problem domains to identify where RL would have the greatest impact. We prioritize collaboration and team culture in addition to being results-oriented, and therefore we always seek researchers and scientists with excellent written and verbal communication skills.

The Role:

• Apply and improve cutting-edge academic research in RL to defense-domain problems (wargames, flight, naval) to train autonomous agents that surpass the performance of existing algorithms
• Build large-scale, real-world autonomous systems by collaborating with STR’s world-class team of engineers and researchers in the machine learning community, our academic partners and our customers
• Analyze data and agent behavior by both replaying simulated scenarios as well as examining real-world data to identify performance improvement strategies
• Interact with national security customers and users to determine and assess unmet needs and evolving requirements
• Transition algorithms to prototype and/or real-time systems demonstrations
• Position may involve travel to customer meetings and/or data collections and experiments

Who You Are:

• BS, MS or PhD degree in Computer Science.
• Academic and/or industry experience training agents to optimize criterion in simulated environments
• Experience with scientific software development and algorithm implementation
• Proficient in Python, Docker, PyTorch, TensorFlow
• Motivated collaborator with internal and external teams

Some Other Relevant Skills You May Have:

• Published work in academic conferences
• Experience with cloud computing (AWS, GCloud, Azure)
• Experience with RLlib, Unity
• Familiar with Agile Development Workflow

STR is a growing technology company with locations near Boston, MA, Arlington, VA, near Dayton, OH, Melbourne, FL, and Carlsbad, CA. We specialize in advanced research and development for defense, intelligence, and national security in: cyber; next generation sensors, radar, sonar, communications, and electronic warfare; and artificial intelligence algorithms and analytics to make sense of the complexity that is exploding around us.