Product Manager, Training Platform

PRODUCT AT BASETEN

Baseten is building a new cloud for AI. Our ambition is to make inference, training, and agent execution part of one coherent platform, and to shape how the next generation of AI products gets built. Product decisions here reach from the economics of GPU infrastructure to the abstractions developers build on every day.

We’re hiring builders and tastemakers with the ambition to help create that future. You should bring a strong point of view about what should exist, the taste to make complex technology feel simple, and the technical judgment to build it alongside engineers, researchers, and designers. You’ll work directly with customers and forward deployed teams, turn emerging capabilities into new possibilities for developers, and carry an idea from early conviction through to production and adoption.

THE ROLE

Help make training and fine-tuning AI models dramatically easier.

Training a model is hard enough. Developers shouldn't have to spend their time wrestling with infrastructure, waiting for GPUs, or figuring out why a job failed.

We're building a training platform that lets AI teams focus on improving their models instead of managing the systems underneath them.

As Product Manager for Training Platform, you'll own the developer experience for running training jobs on Baseten, from environment setup and job submission to scheduling, debugging, checkpointing, and recovery.

You'll work alongside engineers and researchers solving some of the hardest problems in distributed training, helping turn that complexity into a product developers can depend on.

RESPONSIBILITIES

• Own the product experience for training jobs, from setup and submission through execution, monitoring, and recovery

• Make distributed training easier to use, with intuitive workflows for scheduling, GPU availability, logs, checkpoints, and debugging

• Work closely with research and infrastructure teams to support different training workloads, including supervised fine-tuning and reinforcement learning

• Help developers recover from failures, reproduce experiments, and spend less time waiting on infrastructure

• Connect training with the rest of Baseten's platform, including Loops, model evaluation, and inference deployment

• Work directly with customers to understand how they train and adapt models, then turn those insights into product improvements

REQUIREMENTS

• Experience building and shipping AI/ML products or developer platforms

• A technical background in software engineering, ML engineering, research, or equivalent hands-on work

• Direct experience working with teams that train, fine-tune, or adapt machine learning models

• Strong understanding of distributed training concepts, including GPU memory, scheduling, data access, checkpointing, and fault recovery

• Familiarity with the infrastructure needs of supervised fine-tuning and reinforcement learning

• Strong product instincts for developer workflows and the ability to simplify technically complex systems

• Experience partnering with researchers and engineers to take new capabilities from concept to production

NICE TO HAVE

• Hands-on experience running distributed training, building ML developer tools, or working with trainer and sampler systems for reinforcement learning.

WHAT SUCCESS LOOKS LIKE

AI teams can run more experiments, recover from failures faster, and spend more of their time improving models instead of managing compute. Baseten becomes a natural place to train, fine-tune, and deploy custom AI models.

BENEFITS

• Competitive compensation, including meaningful equity

• Paid parental leave

• Fertility and family-building stipend through Carrot

• (U.S. only) Company-facilitated 401(k)