Master Thesis: RL-Accelerated Planning with LLM-Guided Reward Optimization (AI)

Location: Linköping, Sweden; onsite thesis project

Company

hirify.global is a defense and security company developing advanced systems for aeronautics, weapons, command and control, sensors, and underwater applications.

What you will do

• Conduct a literature review on reinforcement learning, motion planning, and LLM-supported reward design.
• Develop a simulation environment for unmanned vehicle motion planning.
• Implement an RL-based planning framework combining learning-based and classical planning methods.
• Integrate an LLM-driven loop for automated reward function design and refinement.
• Run experiments evaluating planning performance, convergence, computational efficiency, and reward optimization.

Requirements

• Be a master’s student completing a 30 HP thesis project.
• Strong theoretical interest in reinforcement learning and motion planning.
• Ability to develop simulation environments and implement experimental machine-learning frameworks.
• Successful completion of security vetting is required; additional citizenship obligations may apply for security-cleared positions.

Nice to have

• Interest in unmanned systems, agentic intelligence, and advanced autonomy.
• Experience with large language models or automated reward engineering.

Culture & Benefits

• Work with the Mission Autonomy group at hirify.global Dynamics.
• Receive support and guidance from experienced engineers and specialists.
• Gain practical experience contributing to autonomy solutions for future unmanned systems.
• Join a technology-focused environment emphasizing innovation and collaboration.