Robotics & AI Intern: Multi-Agent Autonomy & MARL

Laelaps AI Ltd.

Zürich

On-site

CHF 22,000 - 36,000

Full time

4 days ago

Be an early applicant

A complete application in a minute

Tailored resume and cover letter, ready to send.

Get past ATS filters

Benefits offered by this job

Ownership of projects

Ambitious mission in autonomous safety

Strong growth runway

PhD-level mentorship from co-founders

Competitive equity/salary package

Job summary

Laelaps AI Ltd in Zurich is seeking a Software Engineering Intern specializing in Multi-Agent Orchestration to advance the orchestration layer for autonomous agents coordinating across security tasks. You will work on MARL, decentralized planning, and real-world deployment on quadruped robots, with strong mentorship from PhD-level co-founders.

Ideal candidates are in final master's or PhD in CS/Robotics/AI, with experience in reinforcement learning, Python and PyTorch, and hands-on robotics

Qualifications

Currently in the final stages of a Master’s degree or pursuing a PhD in Computer Science, Robotics, AI or a related field.
Demonstrated hands-on experience with Reinforcement Learning.
Experience with robotics simulators such as Isaac Sim, MuJoCo, or Gazebo or custom built simulators.
Solid understanding of reinforcement learning algorithms and concepts.
Ability to independently research, implement, and evaluate algorithms.
Good coding skills in Python, with PyTorch (or JAX).

Responsibilities

Multi-Agent Reinforcement Learning: Design, train, and evaluate MARL algorithms for cooperative multi-robot surveillance and persistent patrolling.
Multi-Robot Coordination: Develop strategies for decentralized decision-making, task allocation, and efficient collaboration between autonomous robots.
Simulation: Improve our simulations to train the policies that can be later transferred to the real world.
Real-World Deployment: Build and test algorithms in simulation environments, with opportunities to deploy them on real quadruped robots.
Algorithm Research & Optimization: Explore and compare learning-based approaches with classical optimization, planning, and heuristic methods.
Performance Evaluation: Develop benchmarks to assess surveillance coverage, patrol efficiency, scalability, and robustness in dynamic environments.