2027 Future Talent Program - AI/ML Computational Toxicology - Intern

Job Description

The Future Talent Program features internships that last up to 12 weeks and will include one or more projects. These opportunities in our Research and Development Division can provide you with great development and a chance to see if we are the right company for your long-term goals. Nonclinical drug safety (NDS) helps advance high-quality drug candidates into development by defining the non–clinical safety and selectivity of lead compounds. NDS employees evaluate the preclinical toxicity of drug development candidates, provide mechanistic understanding of drug-induced toxicity, and assess implications for human safety. Our Computational Toxicology group integrates toxicological expertise, data science, and AI/ML, including emerging Large Language Model (LLM) applications, to improve the safety assessment of drug candidates, accelerate decision-making, and reduce reliance on animal testing. We have multiple openings for highly motivated summer interns to join our AI/ML Computational Toxicology team. The successful candidate will work on innovative projects applying machine learning, natural language processing (NLP), and LLMs to multimodal toxicology datasets, spanning genomics, metabolomics, chemistry and structured in vivo and in vitro data as well as unstructured biomedical literature.

What you will do

Build and evaluate predictive models for toxicity risk and mechanistic hypotheses generation using multimodal pharmaceutical data Develop NLP/LLM pipelines (prompting, fine-tuning, RAG) to mine unstructured reports and literature Prepare dashboards (e.g., Streamlit) and present results to scientists and leadership

Education

Pursuing BS/MS/PhD in Computational Science, Data Science, Bioinformatics, Computational Biology/Chemistry, Computational Linguistics, or related field.

Required Experience and Skills

Proficient in Python; solid with statistics/ML and data wrangling (pandas/SQL) Hands-on with scikit-learn and at least one deep learning framew