Director, Molecular AI & Federated Learning
Eli Lilly and Company
September 03, 2026
On-site
Indianapolis, IN
IT
Job Summary
Director, Molecular AI & Federated Learning—technical leadership for TuneLab uniting privacy-preserving federated learning with generative small-molecule design; leads predictive and generative models to accelerate small-molecule lead optimization and candidate selection (mentorship/vision; no formal people management).
Key Responsibilities
- Set technical direction for federated learning + molecular AI research strategy.
- Mentor scientists/engineers; review methods/code; raise scientific bar.
- Architect federated foundation models for large-scale pretraining on distributed unlabeled/partially labeled data.
- Advance semi-/self-supervised learning for federated constraints (comm bottlenecks, data heterogeneity).
- Develop communication-efficient federated optimization/aggregation (e.g., FedAvg, FedProx, SCAFFOLD).
- Optimize scalability/performance; build simulation environments for benchmarking.
- Architect federated multi-task learning; handle task/feature heterogeneity (personalization/meta-learning; avoid negative transfer).
- Create fine-tuning/adaptation protocols and validation frameworks (per-task metrics; fairness).
- Build multi-task property prediction (ADMET, solubility, permeability, metabolic stability, off-target liabilities).
- Design/deploy generative chemistry models (VAEs, diffusion, flow matching, autoregressive) for de novo design/optimization/scaffold hopping.
- Develop integrated ADMET-driven multi-objective prediction–generation pipelines.
- Ensure synthetic feasibility (reaction-aware generation; retrosynthesis/fragments; collaborate with synthetic chemists).
- Apply XAI; establish benchmarks; publish and maintain reproducible code/version control.
Basic Qualifications
- PhD in CS/Computational Chemistry/Cheminformatics/ML/Computational Biology or related field.
- 5+ years post-PhD ML in drug discovery in biopharma (preference for 8+ years).
Preferred Skills
- Technical leadership without required people-management.
- Generative models for molecular design; multi-task/representation learning.
- Medicinal chemistry + ADMET optimization; federated learning/privacy-preserving ML.
- Publications in top ML venues; GNN/geometric deep learning; organic chemistry/synthetic feasibility.
- RDKit/DeepChem; PyTorch; active learning; uncertainty quantification/XAI; strong cross-disciplinary communication; independent drive.
Location/Travel
- Indianapolis, San Francisco, or Boston; up to 10% travel; attendance at key conferences.
Compensation/Benefits (if applicable)
- Anticipated wage: $177,000–$281,600; bonus eligibility; comprehensive benefits (e.g., 401(k), medical/dental/vision, flexible benefits, life insurance, time off, well-being).
Application Instructions
- Complete the listed accommodations form if you need submission assistance.
Director, Molecular AI & Federated Learning—technical leadership for TuneLab uniting privacy-preserving federated learning with generative small-molecule design; leads predictive and generative models to accelerate small-molecule lead optimization and candidate selection (mentorship/vision; no formal people management).
Key Responsibilities
- Set technical direction for federated learning + molecular AI research strategy.
- Mentor scientists/engineers; review methods/code; raise scientific bar.
- Architect federated foundation models for large-scale pretraining on distributed unlabeled/partially labeled data.
- Advance semi-/self-supervised learning for federated constraints (comm bottlenecks, data heterogeneity).
- Develop communication-efficient federated optimization/aggregation (e.g., FedAvg, FedProx, SCAFFOLD).
- Optimize scalability/performance; build simulation environments for benchmarking.
- Architect federated multi-task learning; handle task/feature heterogeneity (personalization/meta-learning; avoid negative transfer).
- Create fine-tuning/adaptation protocols and validation frameworks (per-task metrics; fairness).
- Build multi-task property prediction (ADMET, solubility, permeability, metabolic stability, off-target liabilities).
- Design/deploy generative chemistry models (VAEs, diffusion, flow matching, autoregressive) for de novo design/optimization/scaffold hopping.
- Develop integrated ADMET-driven multi-objective prediction–generation pipelines.
- Ensure synthetic feasibility (reaction-aware generation; retrosynthesis/fragments; collaborate with synthetic chemists).
- Apply XAI; establish benchmarks; publish and maintain reproducible code/version control.
Basic Qualifications
- PhD in CS/Computational Chemistry/Cheminformatics/ML/Computational Biology or related field.
- 5+ years post-PhD ML in drug discovery in biopharma (preference for 8+ years).
Preferred Skills
- Technical leadership without required people-management.
- Generative models for molecular design; multi-task/representation learning.
- Medicinal chemistry + ADMET optimization; federated learning/privacy-preserving ML.
- Publications in top ML venues; GNN/geometric deep learning; organic chemistry/synthetic feasibility.
- RDKit/DeepChem; PyTorch; active learning; uncertainty quantification/XAI; strong cross-disciplinary communication; independent drive.
Location/Travel
- Indianapolis, San Francisco, or Boston; up to 10% travel; attendance at key conferences.
Compensation/Benefits (if applicable)
- Anticipated wage: $177,000–$281,600; bonus eligibility; comprehensive benefits (e.g., 401(k), medical/dental/vision, flexible benefits, life insurance, time off, well-being).
Application Instructions
- Complete the listed accommodations form if you need submission assistance.