Academic affiliate
The work will entail supervising research in agentic AI and software engineering. The goal is to developing novel agentic AI methods for autonomous code generation and software engineering, focusing on advanced planning algorithms and domain-specific code generation systems. The research will involve designing hierarchical planning architectures, integrating reinforcement learning with large language models, and creating domain adaptation frameworks for specialized applications in robotics, scientific simulation, and embedded systems. The project requires both theoretical algorithm development and empirical evaluation on established and novel benchmarks.
The exact number of weekly hours will be negotiated once a successful candidate has been identified
Agentic AI and Code Generation
- PhD in Computer Science
- Research and teaching experience in AI, robotics and related areas
- Led development and technology transition of many technologies to AI and software systems.
- Strong record of supervision of students in research
- Strong oral and written communication skills
- Experience working with Large Language Models for various applications
Key responsibilities will include but are not limited to:
- Investigating advanced planning and reasoning methods for agentic code generation systems
- Developing domain-specific adaptation frameworks for specialized programming environments
- Designing and implementing novel algorithms for long-horizon planning in software engineering tasks
- Creating benchmarking frameworks and evaluation methodologies for agentic coding systems
- Analyzing performance of different agent architectures across diverse coding domains
- Collaborating with domain experts in robotics, scientific computing, and embedded systems
- Presenting results at internal meetings, conferences, and occasional meetings with external stakeholders
- Ensuring that results, protocols, software, and documentation have been archived or otherwise transmitted to the larger organization