Senior Software Engineer - Model Training

ABOUT BASETEN

Baseten provides the infrastructure, tooling, and expertise needed to bring great AI products to market - fast. Backed by top investors including IVP , Spark Capital , Greylock , and Conviction , we’re trusted by leading AI-driven innovators like Writer , Abridge , Bland , Patreon , Descript, Retool , and Zed to deliver industry-leading performance, security, and reliability for their mission-critical workloads. With our recent $75M Series C funding, we’re growing fast to make AI accessible across all products.

THE ROLE

As a Senior Software Engineer – Model Training at Baseten, you’ll be at the forefront of building the infrastructure that powers large-scale training and fine-tuning of foundation models. You’ll design and implement distributed training systems, optimize GPU utilization, and create scalable pipelines that enable Baseten and our customers to adapt models efficiently and reliably. This role is highly technical and hands-on: you’ll own key components of our training stack, collaborate with product and infra teams to surface customer needs, and push forward the state of the art in scalable training infrastructure.

RESPONSIBILITIES

• Design, build, and maintain distributed training infrastructure for large-scale foundation models

• Implement scalable pipelines for fine-tuning and training across heterogeneous GPU/accelerator clusters

• Optimize training performance through techniques like FSDP, DDP, ZeRO, and mixed precision training

• Contribute to frameworks and tooling that make training workflows efficient, reproducible, and developer-friendly

• Collaborate with cross-functional teams (Product, Forward Deployed Engineering, Inference Infra) to ensure training systems meet real-world requirements

• Research and apply emerging techniques in training efficiency and model adaptation, and productionize them in the Baseten platform

• Participate in code reviews, system design discussions, and technical deep dives to maintain a high engineering bar

REQUIREMENTS

• Bachelor’s degree in Computer Science, Engineering, or related field, or equivalent experience

• 4+ years of experience in software engineering with a focus on ML infrastructure, distributed systems, or ML platform engineering

• Hands-on expertise in distributed training frameworks (FSDP, DDP, ZeRO, or similar) and ML frameworks (PyTorch, Transformers, Lightning, TRL)

• Strong understanding of GPU/accelerator performance optimization and scaling techniques

• Experience designing and operating large-scale systems in production (cloud-native preferred)

• Excellent problem-solving and communication skills, with the ability to work across infrastructure and ML boundaries

NICE TO HAVE

• Experience building APIs, SDKs, or developer tools for ML workflows

• Familiarity with cluster management and scheduling (Kubernetes, Ray, Slurm, etc.)

• Knowledge of parameter-efficient fine-tuning methods (LoRA, QLoRA) and evaluation pipelines

• Contributions to open-source distributed training or ML infra projects

• Experience with cloud environments (AWS, GCP, Azure) and container orchestration

BENEFITS

• Competitive compensation package.

• This is a unique opportunity to be part of a rapidly growing startup in one of the most exciting engineering fields of our era.

• An inclusive and supportive work culture that fosters learning and growth.

• Exposure to a variety of ML startups, offering unparalleled learning and networking opportunities.

Apply now to embark on a rewarding journey in shaping the future of AI! If you are a motivated individual with a passion for machine learning and a desire to be part of a collaborative and forward-thinking team, we would love to hear from you.

At Baseten, we are committed to fostering a diverse and inclusive workplace. We provide equal employment opportunities to all employees and applicants without regard to race, color, religion, gender, sexual orientation, gender identity or expression, national origin, age, genetic information, disability, or veteran status.