Build and improve AI systems for candidate generation, screening, and selecting molecules to test.
Develop energy-based models that combine learned representations with physical and chemical knowledge.
Evaluate models against held-out chemical families, quantum-chemistry calculations, experiments, and prospective tests; track calibration and performance across versions.
Compare predictions with measurements, investigate disagreements, and use findings to improve models.
Summarised from the posting by AI and can be wrong. The full description is the source.
Requirements
Experience
Hands-on experience building ML models for real scientific problems
Experience designing evaluations and experiments that correlate with downstream performance
Strong record of impactful research, shown through publications, patents, open-source software, or methods used in real discovery programs
Hands-on research experience with energy-based models, or methods that combine learning with search and optimization (preferred)
Experience with molecular ML, such as property prediction, generative molecular design, ML interatomic potentials, or quantum-chemistry workflows (e.g. DFT) (preferred)
Experience with closed-loop design using real lab feedback (preferred)
Education
Background in machine learning, computational chemistry, physics, or a related field
PhD, or a Master’s plus relevant industry experience
Skills
Familiarity with energy-based models, diffusion or score-based models, or similar generative approaches
Knowledge of chemistry, or a genuine interest in learning it alongside chemists
Other requirements
Comfort working in an interdisciplinary team
Your match
Sign in and add your resume to see which of this job's requirements it shows.
Saved to your account, encrypted. The AI that reads it doesn't keep it or train on it.
Full job description
From Logical Intelligence’s careers site
At Logical Intelligence, our mission is to accelerate the discovery of materials that make the next generation of technology possible, from light emitters and semiconductors to batteries, electric vehicles, and the chips behind AI and robotics. Large language models alone are not enough to get there. Discovering new materials requires an understanding of chemistry and physics to guide the search, and energy-based models let us build that understanding directly into our discovery loop. We work closely with industry and lab partners to shorten the path from research to real products.
Show full descriptionShow less
About the Role As a Research Scientist on our material discovery team, you will help turn that idea into new molecules. You will build the AI systems that propose and evaluate candidates, and work with our partner labs so that every synthesis and measurement teaches the models something. You will join a highly interdisciplinary team of chemists, physicists, and machine learning scientists and engineers, where the biggest impact comes from working together.
What You’ll Do
Build and improve the AI systems behind our discovery loop, from candidate generation and screening to selecting which molecules to test next.
Develop energy-based models that combine learned representations with physical and chemical knowledge to search chemical space more efficiently.
Own model evaluation, including benchmarks on held-out chemical families, accuracy against both quantum-chemistry calculations and experiment, calibration of uncertainty estimates, prospective tests where the model predicts before the lab measures, and tracking across model versions.
Compare AI predictions with measurements, investigate disagreements, and turn what you learn into concrete model changes, in collaboration with AI researchers and engineers.
Design experiments with our partner labs so that each round of synthesis and testing is as informative as possible.
You Will Thrive in This Role If You Have
A background in machine learning, computational chemistry, physics, or a related field, with a PhD, or a Master’s plus relevant industry experience, and hands-on experience building ML models for real scientific problems.
Familiarity with energy-based models, diffusion or score-based models, or similar generative approaches, and an interest in going deeper.
Knowledge of chemistry, or a genuine interest in learning it alongside our chemists.
Experience designing evaluations and experiments that correlate with downstream performance
A strong record of impactful research, shown through publications, patents, open-source software, or methods used in real discovery programs.
Excellent scientific writing and communication, and comfort working in an interdisciplinary team.
Especially Strong Candidates May Also Have
Hands-on research experience with energy-based models, or with methods that combine learning with search and optimization.
Experience with molecular ML, such as property prediction, generative molecular design, ML interatomic potentials, or quantum-chemistry workflows (e.g. DFT).
Experience with closed-loop design using real lab feedback