SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Preference Model is building automated ML research engineering, focusing on creating RL training environments that reflect real-world complexity to enable frontier models to perform better on practical tasks. The founding team includes former members of Anthropic's data team who built infrastructure and datasets for Claude.
As a Senior ML Infrastructure Engineer, you will design, build, and scale the compute, scheduling, and data infrastructure powering post-training research on in-house RL environments. You'll develop and maintain core ML framework primitives and internal tooling that researchers depend on daily, accelerating reproducible experimentation and reducing iteration time. You'll build evaluation and benchmarking infrastructure, monitoring, logging, and debugging tooling, plus automated testing and deployment systems to catch failures early and maintain reliability at scale. You'll partner directly with Research Engineers to translate research needs into infrastructure requirements and ship fast based on their feedback.
The role sits at the intersection of production systems engineering and frontier AI research, requiring you to balance infrastructure rigor with the pace of a fast-moving research environment.
REQUIREMENTS:
- 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 inference libraries (vLLM, SGLang, Megatron)
- Experience with RL training frameworks (Slime, veRL, Ray Train, SkyRL)
- Significant experience and understanding of distributed systems principles, with hands-on cloud platform experience (AWS, GCP) and container orchestration (Kubernetes) for high-throughput, low-latency workloads
- Experience with data engineering tools and building robust, scalable data pipelines
- Proficiency in core ML frameworks (PyTorch or JAX)
- Ability to balance production rigor with fast-moving research pace and communicate infrastructure tradeoffs clearly to non-infrastructure specialists
