Preference Model is building the next generation of training data to power the future of AI. Backed by a16z and South Park Commons.
About the role
Frontier research moves only as fast as its infrastructure permits. Building solid infrastructure is foundational to our mission of pushing self-directed learning as far as it can go.
What they're looking for
- Strong software engineering fundamentals and hands-on experience building production-grade LLM inference and training infrastructure (ideally from the ground up)
- Experience building LLM training/inference internals such as transformers, distributed training, and working on inference libraries like vLLM, SGLang, Megatron
- Experience working on RL training frameworks like Slime, veRL, Ray Train, SkyRL
- Significant experience and understanding of distributed systems principles, and have hands-on experience with cloud platforms (AWS, GCP) and container orchestration (Kubernetes), building systems for high-throughput, low-latency workloads
- Have experience with data engineering tools and building robust, scalable data pipelines
- Proficiency in core ML frameworks such as PyTorch or JAX
More about this role
Preference Model is building automated ML research engineering.
Existing frontier models are brittle when applied to real-world ML tasks. The present bottleneck is the lack of high-quality RL training environments. Our first step is to build RL environments that reflect real-world complexity, with diverse tasks and robust reward functions.
Our founding team has previous experience on Anthropic’s data team building data infrastructure, and datasets behind Claude. We are partnering with leading AI labs to push AI closer to achieving its transformative potential.
Frontier research moves only as fast as its infrastructure permits. Building solid infrastructure is foundational to our mission of pushing self-directed learning as far as it can go.
We are looking for Senior ML Infrastructure Engineers to build the infrastructure and systems that power the frontier of post-training on large language models . This role involves building scalable infrastructure to enable high-throughput systems and shape how our research is run, bringing us closer to models that can train themselves on what they aren't yet good at.
Design, build, and scale the compute, scheduling, and data infrastructure...
Browse similar: AI jobs · AI startup jobs · Startup jobs · San Francisco Bay Area