Senior ML Engineer — Production Systems & AI Evaluation

Snorkel AI; New York (NY); On-site; USD 180,000 - 240,000; Full time

Job summary

Snorkel AI seeks an experienced engineer to develop and scale frontier AI data generation and evaluation methods, focusing on faster, cheaper, and more rigorous processes. The role involves engineering production ML workloads, model routing, and fine-tuning, turning research prototypes into reusable production components for engineers and researchers.

Candidates should have 5+ years in production ML, strong Python, and a solid statistics background, with proven ability to design evaluations and

Qualifications

• 5+ years building production ML or software systems.
• Hands-on experience with production LLM/ML workloads.
• Strong Python and software engineering fundamentals.
• Rigorous grounding in statistics and experimentation.
• Experience designing evaluations and interpreting results.
• Fine-tuning models, evaluation platforms, agentic workloads, and research-to-production experience preferred.

Responsibilities

• Develop and scale methods that make frontier AI data generation and evaluation faster, less expensive, and more rigorous, including efficient agent evaluations, model routing, fine-tuned models, and predictive task difficulty.
• Establish measurement standards and turn research prototypes into reusable production components for engineers and researchers.

Skills

Production Machine Learning

Python

Statistics

Experiment Design

Hypothesis Testing

Sampling

Confidence Intervals

LLM Evaluation

Model Routing

Fine-Tuning

Open-Weight Models

Agentic Workloads

Software Engineering

System Design

Testing

Research-to-Production

Job description

Snorkel AI seeks an experienced engineer to develop and scale frontier AI data generation and evaluation methods, focusing on faster, cheaper, and more rigorous processes. The role involves engineering production ML workloads, model routing, and fine-tuning, turning research prototypes into reusable production components for engineers and researchers.

Candidates should have 5+ years in production ML, strong Python, and a solid statistics background, with proven ability to design evaluations and