SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Lila Sciences is seeking a Computational Scientist to develop machine learning models and workflows that accelerate discovery in polymeric and soft material systems. The role focuses on solids and viscoelastic materials including polymers, elastomers, gels, hot-melt adhesives, composites, powders, films, and semi-solids.
You will bridge domain expertise in polymer science, soft matter physics, rheology, or formulation science with machine learning execution. Core responsibilities include developing structure-property models that connect formulation choices, processing history, material morphology, and end-use performance. You will build representations incorporating molecular descriptors, simulation outputs, and mechanistic constraints; design active learning workflows aligned with physical formulation workcell throughput; and create tools to help scientists interpret material data and prioritize formulation decisions.
Key technical areas span melt processing, mechanical performance, thermal transitions, processing windows, crystallinity, cross-link density, cure kinetics, and formulation-to-processing-to-property relationships. You will work with experimental, simulation, rheological, thermal, and mechanical datasets to train and improve models. Strong collaboration with experimental teams is essential to align models with measurement workflows and practical formulation development needs.
Required qualifications include experience applying machine learning to materials or formulation problems, domain expertise in polymer science or related fields, strong Python and modern ML framework skills, and ability to evaluate models using scientific datasets. A PhD in chemical engineering, materials science, physics, applied mathematics, or related field is expected, or a master's degree with equivalent relevant experience. Bonus experience includes working with polymer experimental data, modeling structure-property relationships for viscoelastic materials, incorporating molecular descriptors into ML workflows, and background in active learning systems that close loops between models and high-throughput experimentation.
