Senior ML Engineer, Production LLMs & RAG | Hybrid Seattle

Fuel Talent; Seattle (WA); Hybrid; USD 160,000 - 230,000; Full time

Job summary

Fuel Talent is partnering with a mission-driven startup in Seattle to hire a Sr. Machine Learning Engineer. You will own model performance in production, fine-tune foundation models, and build retrieval-augmented systems for enterprise use.

This hybrid role offers a base salary of $160K–$230K plus equity and benefits. You will work with Python, PyTorch, and modern ML tooling, shaping experiments from data prep to deployment while balancing latency and cost.

Qualifications

• 4+ years of ML or applied AI experience with production models.
• Strong Python and hands-on experience with PyTorch or Hugging Face.
• Practical experience with LLMs: fine-tuning, RAG, prompts, agents, or evals.
• Experience designing evaluation pipelines to measure model quality.
• Familiarity with containerization and cloud-based deployment.

Responsibilities

• Fine-tune, adapt, and optimize LLMs and foundation models for domain-specific use.
• Design and build agentic workflows and retrieval-augmented systems from prototype to production.
• Build rigorous evaluation and benchmarking systems to prove improvements.
• Own model performance end-to-end: data prep, experiments, deployment, monitoring.
• Collaborate with engineering and product to translate business problems into ML solutions.
• Ship models reliably with latency and cost considerations.

Skills

Python

LLMs

Model evaluation

Production ML

Latency optimization

Tools

PyTorch

Hugging Face

Docker

Kubernetes

Terraform

Job description

Fuel Talent is partnering with a mission-driven startup in Seattle to hire a Sr. Machine Learning Engineer. You will own model performance in production, fine-tune foundation models, and build retrieval-augmented systems for enterprise use.

This hybrid role offers a base salary of $160K–$230K plus equity and benefits. You will work with Python, PyTorch, and modern ML tooling, shaping experiments from data prep to deployment while balancing latency and cost.