Overview

Looking for a Lead Engineer to join a Tech Bio company to lead the implementation of novel Generative AI technologies

Responsibilities

• Lead large scale model training
  
  
  
  • Design, implement, and optimize distributed training (DDP/FSDP), mixed precision, checkpointing/restarts, and experiment throughput/cost on multi-GPU/TPU clusters.
  
  
  • Build reliable training pipelines and artifacts (data → train → eval) with strong failure recovery and reproducibility controls.
  
  
  
  


• Implement state-of-the-art models for research enablement
  
  
  
  • Implement production-grade Transformers, Diffusion, Flow-Matching variants; build training loops, eval harnesses, and ablations.
  
  
  • Profile and optimize memory/latency
  
  
  • Establish CI/CD/CT for ML, model versioning, lineage and promotion/rollback using a model registry; enforce experiment tracking and data/model provenance.
  
  
  • Partner with the engineering team to serve the models into products: Define serving paths (batch/online), observability (drift/perf/SLA alerts), and safe rollout/rollback procedures
  
  
  
  


• Technical leadership (IC)
  
  
  
  • Set coding standards, reviews, documentation, and reproducibility practices within the R&D
  
  
  
  

Your profile

• Track record shipping large-scale ML systems (6–10+ years): deep Python, strong software engineering (testing, profiling, debugging), and hands-on distributed training on multi-GPU clusters.


• Proven implementation expertise in at least two of: Transformers, Diffusion Models, Flow Matching—from reference-quality code to optimized training at scale.


• MLOps proficiency: CI/CD/CT for ML, experiment tracking, artifact and data lineage, model registry workflows, containerization/orchestration (Docker/Kubernetes), and cloud.


• Inference & serving: experience packaging models for batch/online use, monitoring performance & drift, and executing safe rollouts.


• Biology experience: comfort with perturbation/omics datasets and partnering with scientists; familiarity with causal objectives and evals.


• Leadership & collaboration: clear communication, design-doc discipline, and demonstrated mentorship as a tech lead