← All roles

Member of Technical Staff, Post-Training & Applied Research

SF Tensor· SF Tensor
Apply on company site
San Francisco$275k - $315kFull-timePosted 1mo ago

About the role

Role within SF Tensor focusing on end-to-end post-training and applied research to accelerate AI compute and model deployment. Owns post-training pipeline from data curation to deployment, with emphasis on SFT, RL, distillation, and customer-focused reward design.

Responsibilities

  • Own the post-training pipeline end-to-end: data curation, SFT, preference optimization, RL, evals, distillation and finally deployment
  • Design reward functions and RL looks with customer domain experts
  • Build an eval harness trustworthy enough to make ship/no-ship decisions within a short window
  • Structure and generate datasets, including synthetic data pipelines, from customer data
  • Distill specialist models down into smaller models
  • Drive time-to-model by identifying and removing the critical path in experiments
  • Embed with customers as a forward-deployed researcher and hand pipeline over when ready to scale

Qualifications

  • Shipped post-trained models into production and can discuss tradeoffs
  • Hands-on depth across SFT and RL (DPO, GRPO, PPO or similar)
  • Ability to judge evaluation: what to measure, interpretation of results, detecting misleading numbers
  • Experience owning data curation, filtering, labeling workflows and synthetic data generation
  • Proficiency in PyTorch or JAX
  • Willingness to work with customers' domain experts to translate intuition into reward functions

Skills

Required:Post-training pipelineReinforcement LearningEvaluation & Reward DesignData Curation & Synthetic DataPyTorch or JAX

Benefits

  • Meaningful equity and benefits (as per description)
  • relocation assistance (where possible)
  • US-based compensation with base salary and equity
Last updated Oct 7, 2026. Interested? Apply on the company site.
SF Tensor logo

About SF Tensor

Infrastructure for AI labs to focus on research.

AI researchers should be pushing the boundaries of what's possible with new architectures and training methods. Instead, they waste weeks configuring cloud infrastructure, debugging distributed systems, and optimizing their GPU code. We know because we lived it: While training our own models across thousands of GPUs earlier this year, we spent more time fighting our infrastructure than doing actual research. That why we're building two things. First, Elastic Cloud: a managed platform that automatically finds the cheapest GPUs across all providers, handles spot instance preemption, and cuts compute costs by up to 80%. Second, automatic kernel optimization that makes training code run faster by modeling hardware topology, often beating hand-tuned implementations. The problem is that getting high performance across different hardware is genuinely hard. NVIDIA's CUDA moat exists because writing fast kernels requires deep expertise. Most teams either accept vendor lock-in or hire expensive kernel engineers. Our goal is to break the CUDA moat. The compute bottleneck is the biggest constraint on AI progress. NVIDIA can't manufacture enough GPUs, and their monopoly keeps prices astronomical. Meanwhile, AMD, Google, and Amazon are shipping capable alternative hardware that nobody uses because the software is too hard. We're breaking that moat. If we succeed, anyone will be able to train state-of-the-art models without thinking past their PyTorch code.

machine-learningopen-sourcecloud-computingai
Company profileWebsite

More open roles at SF Tensor