Senior Product Engineer - Training Platform

ABOUT BASETEN

Baseten powers mission-critical inference for the world's most dynamic AI companies, like Cursor, Notion, Open Evidence, Abridge, Clay, Gamma and Writer. By uniting applied AI research, flexible infrastructure, and seamless developer tooling, we enable companies operating at the frontier of AI to bring cutting-edge models into production. We're growing quickly and recently raised our $300M Series E, backed by investors including BOND, IVP, Spark Capital, Greylock, and Conviction.

Join us and help build the platform engineers turn to to ship AI products.

We’re looking for a customer-obsessed software engineer to come ship with us. You’ll own features like multi-node training and products like serverless reinforcement learning (RL) from conception to MVP (and from MVP to GA!). You’ll work through the stack, architecting solutions from API and UI down to our infrastructure layer. You’ll fine tune models yourself to develop an understanding of user workflows.

You’ll work closely with research engineers leveraging state-of-the-art training techniques to build experiences that accelerate model development and solve for real pain points. If you’re excited to dive deep into the training, let’s talk!

Take a look at what we’ve built so far:

• Overview of the product so far
• Training docs overview
• Story of the Training product
• Research we've done

• Checkpointing Pipeline:
  Our checkpointing pipeline starts with automated checkpointing, a feature that ensures that versions of models created during training are automatically backed up to the cloud. Users are able to then deploy checkpoints seamlessly into inference servers, providing point-and-click integrations into inference frameworks like vLLM and Baseten’s Inference Stack. This enables customers to quickly evaluate the performance of their checkpoints with real traffic.
• Multinode training:
  Multinode training enables customers to easily run training jobs across multiple compute nodes, enabling users to train large models like GLM 4.7 and Deep Seek. We’ve built deeply at the Kubernetes layer to ensure that scheduling, startup, inter-node communication, and shutdown happen seamlessly under the hood and as the user expects.
• Training DX:
  Customers come to train on Baseten because it helps them get to value fast. To do this, we ensure that the features we ship aren’t just fast, but are easy to iterate with. We enhanced Baseten’s metrics from pod-level GPU summaries to per-GPU and per-Node. We’ve built a CLI experience that caters to terminal users, and UI experiences that enable user to seamlessly manage their training jobs.

• Iterate like crazy
• Design ergonomic APIs and abstractions to model complex resources and life cycles
• Work throughout the stack (API layer, backend and database implementation, infra layer; frontend is a plus) to implement features.
• Fine-tune and deploy models to develop intuition around training workflows.
• Partner closely with model developers and world-class research engineers to understand the requirements and pain points of post-training workflows.
• Drive long-term improvements to improve reliability of systems and velocity of development
• Fix bugs & resolve customer issues with urgency

• 5+ years experience building software applications
• Deep knowledge of the web stack, databases, and distributed systems
• Experience developing developer tooling or infrastructure products for external or internal users.
• Good taste in product, particularly developer-oriented tools
• Interest in ML/AI infrastructure and willingness to learn
• Driven by high agency and ownership
• Strong communication skills with the ability to bridge technical depth and business needs

• Experience launching features and products through different release cycles (MVP, Beta, GA, etc.)
• Experience with model development methods and paradigms, like Supervised Fine-Tuning, Reinforcement Learning, Synthetic Data Generation, LoRA, Full Finetunes, etc.
• Familiarity or experience with the open source training stack and frameworks (NCCL, PyTorch, Megatron, NemoRL, VeRL, Axolotl, HF Trainer) and distributed training techniques…