Associate Director, Intelligence Systems Lab
Bristol Myers Squibb
4 hours ago
Remote friendly (Seattle, WA)
United States
IT
Responsibilities:
- Lead research on AI agent intelligence (reasoning, planning, memory, generalization) and architect agent cognition frameworks (abstraction, goal decomposition, adaptive planning) for multi-agent systems.
- Establish benchmarking/evaluation science: define success criteria, build test harnesses/evaluation datasets, ensure rigor/reproducibility, and set strategy for measuring/validating agent intelligence.
- Design/develop agentic AI systems (e.g., Strands, LangGraph, DSPy) including RAG applications, knowledge graphs, conversational AI, and autonomous workflows/decision systems.
- Prototype and innovate: translate research into production-ready demonstrations; prioritize work using ROI-first evaluation; document findings.
- Collaborate cross-functionally to translate research into business value and enable deployment; mentor team and establish governance/best practices.
Qualifications (required/preferred):
- BA/BS required (quantitative field); PhD preferred.
- 5+ years hands-on data science/ML/AI research or development (leading teams preferred; pharma/life sciences preferred).
- Proficient in Python; ML frameworks (e.g., Scikit-Learn/TensorFlow/PyTorch); agentic frameworks (e.g., Strands/LangGraph/AutoGen/DSPy/CrewAI).
- Experience with LLMs, prompt engineering, RAG; benchmarking/evaluation methods; SQL; AWS/Azure/GCP; Git, MLOps, containers; experimental design/statistics.
- Excellent communication/presentation; conference/publishing preferred.
Benefits/Compensation (explicit): Health, wellbeing, 401(k), disability/life/insurance; paid time off; salary range varies by location (e.g., $167,540–$223,314).
Application instruction: Apply even if the resume doesn’t perfectly match.
- Lead research on AI agent intelligence (reasoning, planning, memory, generalization) and architect agent cognition frameworks (abstraction, goal decomposition, adaptive planning) for multi-agent systems.
- Establish benchmarking/evaluation science: define success criteria, build test harnesses/evaluation datasets, ensure rigor/reproducibility, and set strategy for measuring/validating agent intelligence.
- Design/develop agentic AI systems (e.g., Strands, LangGraph, DSPy) including RAG applications, knowledge graphs, conversational AI, and autonomous workflows/decision systems.
- Prototype and innovate: translate research into production-ready demonstrations; prioritize work using ROI-first evaluation; document findings.
- Collaborate cross-functionally to translate research into business value and enable deployment; mentor team and establish governance/best practices.
Qualifications (required/preferred):
- BA/BS required (quantitative field); PhD preferred.
- 5+ years hands-on data science/ML/AI research or development (leading teams preferred; pharma/life sciences preferred).
- Proficient in Python; ML frameworks (e.g., Scikit-Learn/TensorFlow/PyTorch); agentic frameworks (e.g., Strands/LangGraph/AutoGen/DSPy/CrewAI).
- Experience with LLMs, prompt engineering, RAG; benchmarking/evaluation methods; SQL; AWS/Azure/GCP; Git, MLOps, containers; experimental design/statistics.
- Excellent communication/presentation; conference/publishing preferred.
Benefits/Compensation (explicit): Health, wellbeing, 401(k), disability/life/insurance; paid time off; salary range varies by location (e.g., $167,540–$223,314).
Application instruction: Apply even if the resume doesn’t perfectly match.