Advisor - Agent Research

Eli Lilly Eli Lilly · Pharma · Indianapolis, IN +5

Seeking a scientist-engineer hybrid to deploy AI-driven discovery platforms using foundation models, multi-agent systems, and robotics in drug discovery workflows. The role involves translating scientific workflows into agentic systems, integrating LLM reasoning with domain tools, and supporting model deployment and inference services.

What you'd actually do

  1. Partner with chemists and biologists to translate scientific workflows into agentic systems
  2. Deploy and integrate Agentic AI system into active research programs
  3. Design and implement cloud-native data pipelines connecting lab instruments, databases, and AI models
  4. Support model deployment, inference services, and experiment tracking (e.g., MLflow)
  5. Integrate LLM reasoning with domain tools (RDKit, molecular graph ML, ELN/LIMS APIs, instrument drivers) to build composite agents that plan, simulate, and execute DMTA tasks

Skills

Required

  • Python
  • ML/Deep Learning frameworks (PyTorch, Tensorflow, JAX, HuggingFace)
  • building agentic AI systems (e.g., LangChain, OpenAI Agents SDK)
  • designing and shipping end-end systems in cloud environments
  • DevOps/engineering skills: version control (git), containerization (docker, kubernetes), GitOps + CI/CD practices, data systems (Redis, SQL/NoSQL), unit testing, frontend (streamlit, flask)
  • cloud-native (AWS/Azure) pipeline architectures including Nextflow, Argo on Kubernetes
  • MLOps, including model versioning, data versioning, and continuous integration/continuous deployment for ML systems

Nice to have

  • LLM post-training, fine-tuning, or RLHF
  • Experience mentoring and guiding junior researchers or engineers

What the JD emphasized

  • Measurable reduction in DMTA turnaround through autonomous planning and execution
  • Seamless transition from prototype to production-deployed AI systems
  • Demonstrable research experience, evidenced by contributions to projects, and ideally through publications in relevant ML/NLP venues (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP)

Other signals

  • deploying AI-driven discovery platforms
  • integrating foundation models, multi-agent systems, and robotics
  • translating scientific workflows into agentic systems
  • integrating LLM reasoning with domain tools