Location: Boston Massachusetts
Type: Full time
I am representing a deep tech organisation building an AI powered materials discovery platform. Their system integrates advanced model architecture with computational chemistry and large scale simulation. The company has strong financial backing, over 200 million USD in liquidity, and partnerships with leading automotive and energy companies, giving real-world impact to its technology.
About the platform
At the heart of the company’s work is a high-dimensional system that maps and predicts molecular behaviour at scale. It combines physics-informed models, learned representations, and automated simulations to accelerate discovery of next-generation battery and electrolyte materials. This system also supports agentic reasoning, interpretable predictions, and causal insights, allowing the platform to guide scientific exploration autonomously.
What you will work on
- Design and develop foundation models and agentic AI systems for scientific applications
- Apply multimodal reasoning and causal inference to molecular and materials datasets
- Build interpretable models with next-generation explainability tools for scientific ML
- Develop representation engineering approaches to improve model generalisation and accuracy
- Work closely with chemists, physicists, and engineers to translate scientific challenges into ML solutions
- Contribute to experiments that validate and benchmark AI predictions in real-world materials discovery
- Mentor and collaborate with team members across levels to share best practices in AI research
- Strong background in ML, foundation model design, or AI research applied to scientific domains
- Experience in molecular simulation, materials science, or physics-informed ML
- Knowledge of agentic AI, causal reasoning, multimodal reasoning, or interpretable ML techniques
- Familiarity with large datasets in chemistry, materials, or battery sciences
- Ability to communicate complex AI concepts to scientific and technical teams
- Research mindset with curiosity for building models that can reason and act autonomously
