onsite
AI Agent Scientist
AI Agent Scientist
The AI Agent Scientist will drive the development of advanced AI systems for materials design, leveraging LLM architectures and multi-agent systems to achieve complex, multi-objective goals. This role involves synthesizing high-quality data, developing robust evaluation methods, and collaborating with materials science teams, while also contributing to the AI research community and mentoring junior colleagues.
About the role
About the Role
As an AI Agent Scientist, you will be instrumental in driving the development of sophisticated systems for materials design. This role requires a blend of creativity, technical expertise, and a passion for pushing the boundaries of AI in scientific discovery.
Responsibilities
- Drive the development of a system that can pursue complex, multi-objective goals in materials design with limited direct supervision, leveraging agents and multi-agents that can interact with third-party and in-house integrations, including APIs, code interpreters, and hardware components.
- Leverage various LLM architectures and models to balance reasoning and continuous adaptation towards intelligent decision-making.
- Synthesize large-scale, high-quality (multi-modal) data through methods such as rewriting, augmentation, and generation to improve the abilities of models in various stages.
- Work in collaboration with various teams, in particular the experimental materials science team.
- Develop and apply reliable evaluation methods to analyze model performance at different stages, uncover the mechanisms and origins of their capabilities, and leverage these insights to enhance model development.
- Contribute to the AI and materials science research community through publications and active engagement in top-tier AI/ML venues like NeurIPS, ICML, ICLR, and our company blog, alongside active participation in conferences and workshops.
- The ability to tackle challenging problems with new and different ideas, creativity and contrarian thinking.
- Mentor and guide junior team members and interns, promoting an environment of continuous learning and innovation.
Skills
Llm ArchitecturesModel Evaluationdata augmentationmulti modal datamaterials designAiML