Tech Lead/Manager – Model Training

Baseten

full-time

Posted on: 9/4/2025

Location: California • 🇺🇸 United States

✨ AI Apply

💰 $250,000 - $300,000 per year

Senior

AWSAzureCloudDistributed SystemsGoogle Cloud PlatformKubernetesRaySpark

About the role

Lead, mentor, and grow a team of engineers building Baseten’s training infrastructure
Define and drive the technical strategy for large-scale training systems, with a focus on scalability, reliability, and efficiency
Architect and optimize distributed training pipelines across heterogeneous GPU/accelerator environments
Balance hands-on contributions (system design, code reviews, prototyping) with people leadership and career development
Establish best practices for training workflows, distributed systems design, and high-performance model evaluation
Collaborate with Product and Platform Engineering to translate customer and internal needs into reusable infrastructure and APIs
Develop processes that ensure consistent, reliable, and on-time delivery of high-quality systems
Stay ahead of the curve on advancements in training efficiency (FSDP, ZeRO, parameter-efficient training, hardware-aware scheduling) and bring them into production

Bachelor’s degree in Computer Science, Engineering, or related field, or equivalent experience
5+ years of experience in ML infrastructure, distributed systems, or ML platform engineering, including 2+ years in a tech lead or manager role
Strong expertise in distributed training frameworks and orchestration (FSDP, DDP, ZeRO, Ray, Kubernetes, Slurm, or similar)
Hands-on experience building or scaling training infrastructure for LLMs or other foundation models
Deep understanding of GPU/accelerator hardware utilization, mixed precision training, and scaling efficiency
Proven ability to lead and mentor technical teams while delivering complex infrastructure projects
Excellent communication skills, with the ability to bridge technical depth and business needs
Nice to have: Experience with multi-tenant, production-grade ML platforms
Nice to have: Familiarity with cluster management, GPU scheduling, or elastic resource scaling
Nice to have: Knowledge of advanced model adaptation techniques (LoRA, QLoRA, RLHF, DPO)
Nice to have: Contributions to open-source distributed training or ML infrastructure projects
Nice to have: Experience building developer-friendly APIs or SDKs for ML workflows
Nice to have: Cloud-native infrastructure experience (AWS, GCP, Azure, containerization, orchestration)