Principal AI Engineer (GenAI) - Molecular Discovery

Role Summary

Own the strategy and delivery of LLM-native applications, agentic predictive-model workflows, and insight-driven analytics platforms that accelerate large and small-molecule invention. Translate scientific objectives into intuitive software products and robust model-ops practices that help chemists, protein engineers, and data scientists iterate faster, uncover deeper insights, and make better decisions.

Responsibilities

• Using Agentic LLM frameworks, build scientifically grounded conversational analytics, automated reports, and copilot workflows with complex scientific datasets and tools.
• Deliver full-stack applications, React/Next.js front-ends with Python/FastAPI and GraphQL services that surface models and analytics at scale.
• Champion predictive-model use-cases across small and large molecule discovery (e.g., property prediction, sequence optimization, generative design).
• Familiarity with platforms that orchestrate cutting-edge structure- and sequence-prediction toolkits (RDKit, OpenEye, Schrödinger LiveDesign, AlphaFold) for CADD, sequence design and developability assessment.
• Stand up automated pipelines for data curation, experiment tracking, CI/CD, and governed model release (PyTorch/TensorFlow + MLflow/Kubeflow/SageMaker + GitHub Actions).
• Package and deploy predictive applications and model endpoints to cloud-native MLOps or on-prem containers for scalable inference and performant access.
• Codify reusable templates, inner-source libraries, and design systems that cut feature time-to-value.
• Mentor a cross-disciplinary team of full-stack, data and AI engineers; foster best practices in code quality, documentation, and UX research.
• Partner with discovery leads, IT operations, and external vendors to align technical backlogs with portfolio milestones and data-quality standards.
• Influence budgeting and make-vs-buy decisions for AI tooling and platform enhancements.

Qualifications

• Required: Agentic LLM engineering depth - Demonstrated success building Agentic LLM workflows and LLM-native applications; fine-tuning and deploying LLMs, diffusion models, or vision transformers for scientific or operational use-cases.
• Required: Discovery & Molecular Tooling Context - 5+ years with relevant advanced degree building/supporting platforms and tools for computational compound design and protein engineering workflows (Schrödinger, OpenEye, AlphaFold); fluent in SAR analysis, sequence/structure predictions, and assay lifecycles.
• Required: Modern MLOps - IaC (Terraform/CloudFormation), automated testing, secrets management, continuous model evaluation, lineage tracking.
• Required: Influence & communication - lead architecture reviews, map tech choices to scientific KPIs, mentor cross-functional teams, and guide roadmap workshops with executives and bench scientists alike.

Skills

• Contributions to open-source molecular-design projects.
• Advanced Python & React; shipped production apps that integrate APIs, scale model inference, and manage complex research datasets.
• Comfortable packaging and operating applications/models on Kubernetes/EKS, serverless FaaS, or on-prem containers.
• Knowledge of GPU runtime tuning or Triton-based multi-model serving.
• Experience crafting templates or inner-source libraries that accelerate team velocity.
• Cloud-architect certifications (AWS Pro, Azure Expert, etc.).
• Multi-cloud deployment mastery (AWS, Azure, GCP).

Education

• M.S. or Ph.D. in Computer Science, Machine Learning, Computational Chemistry/Biology, or related field; Cloud-architect certification a plus.