ML Engineer: Production-Scale LLMs & AI Infra

Harrison Clarke; San Francisco (CA); On-site; USD 150,000 - 190,000; Full time

Job summary

Harrison Clarke is seeking a technically talented engineer to join a well-funded AI startup in San Francisco, building cutting-edge ML systems at the intersection of data pipelines, model training, and production inference. You will own the full ML lifecycle—from data generation and training to deployment and monitoring, collaborating with researchers to turn ideas into reliable production systems.

A strong foundation in Python and ML frameworks, plus knowledge of transformers, is required, with

Qualifications

• Bachelor's or Master's in CS or a related technical discipline.
• Strong Python programming skills and experience with modern ML frameworks.
• Solid understanding of transformer architectures and large language models.
• Experience building production-quality ML systems.
• Comfortable working across both research and engineering environments.
• Strong software engineering fundamentals and systems thinking.

Responsibilities

• Building scalable data pipelines to collect, process, and generate large synthetic datasets for ML.
• Profiling and optimizing model training and inference performance.
• Deploying and maintaining high-throughput inference systems for LLMs.
• Working closely with researchers to translate new ideas into reliable production systems.
• Building tooling that supports the ML development lifecycle from experimentation through deployment.
• Contributing to technical research, experimentation, and engineering best practices.

Skills

Python programming

ML frameworks

Transformer architectures

Production ML systems

Software engineering fundamentals

Research & engineering collaboration

Education

Bachelors/Masters in CS or related

Tools

DeepSpeed

FSDP

Ray

Spark

Beam

Job description

Harrison Clarke is seeking a technically talented engineer to join a well-funded AI startup in San Francisco, building cutting-edge ML systems at the intersection of data pipelines, model training, and production inference. You will own the full ML lifecycle—from data generation and training to deployment and monitoring, collaborating with researchers to turn ideas into reliable production systems.

A strong foundation in Python and ML frameworks, plus knowledge of transformers, is required, with