About the Role

This is a 12-week PhD/MSc internship at Microsoft Teams CMD Labs to research and develop live proactive meeting agents using reinforcement learning. The intern will train LLMs in simulated environments, deploy them in live Teams scenarios, and publish results at leading AI conferences. The role is based in Cambridge, UK.

Key Responsibilities

• Conduct experiments and train LLMs for proactive behavior in live meeting scenarios
• Create and validate metrics for candidate models
• Collaborate with researchers and engineers to leverage existing assets and datasets
• Ensure results can be integrated back into the product
• Present results internally and write up work for publication in leading AI conferences

Required Skills & Qualifications

Must Have:

• Currently enrolled in a PhD program or published candidate in MSc program in Computer Science, Electrical/Computer Engineering, Statistics, or related field
• Practical experience training/fine-tuning LLMs using Reinforcement Learning in synthetic training environments
• Practical Python coding experience with PyTorch or TensorFlow or similar framework
• Excellent analytical, coding, communication, and collaborative skills

Nice to Have:

• Field of research and publications directly related to language modeling, LLMs, Reinforcement learning
• Experience in model quantization, pruning or distillation
• Experience working in the domain of live chat or meeting scenarios

Benefits & Perks

• Industry leading healthcare