AFRL — Reinforcement Learning Research

Computer Scientist at the Air Force Research Laboratory (AFRL), Rome, NY.

March 2024 – July 2025 · Active Secret Clearance

Research Focus

Developed reinforcement learning agents for autonomous decision-making in complex environments.

Technical Contributions

• Transformer-based RL agents: Designed and trained agents using transformer architectures for sequential decision-making
• Graph-DQN: Implemented Deep Q-Networks leveraging graph neural networks for structured environment reasoning
• Custom environments: Built OpenAI Gym environments for training and evaluating agents across mission-relevant scenarios
• Multi-agent systems: Researched and prototyped multiple agent architectures

Technologies

Python, PyTorch, OpenAI Gym, Transformers, Graph Neural Networks, Deep Q-Networks