Staff Engineer, ML Evaluation & Model Improvement

Sierra Ventures; San Francisco (CA); On-site; USD 150,000 - 210,000; Full time

Benefits offered by this job

Top-tier compensation

Stock options

Health & wellness

Job summary

Reflection, a research lab, seeks to advance open and accessible intelligence. You will conduct rigorous comparative analyses, build evaluation systems, and develop frameworks that capture reasoning and alignment advances.

You will collaborate with pre-training, post-training, and applied teams to translate insights into model improvements. The role emphasizes high agency in a fast-paced startup environment, with the chance to push measurable boundaries and contribute to open foundational

Qualifications

• Strong statistical analysis and experimental design skills to rigorously measure model improvements.
• Familiarity with LLM evaluation methodologies: static benchmarks, human preference evals, and/or agentic tasks.
• High agency and thrive in a fast-paced startup environment; bias for impact over process.
• Excited to work in a new frontier lab, defining how we measure and accelerate progress toward more capable models.
• Collaborative, detail-oriented, and motivated by building the feedback loops that make models truly improve.

Responsibilities

• Conduct critical comparative analysis to advance our understanding of model capabilities.
• Build and refine evaluation systems and processes that create tight feedback loops between data, evals, and model behavior.
• Develop generalizable evaluation frameworks that capture what matters for reasoning, alignment, and usefulness.

Skills

Statistical analysis

Experimental design

LLM evaluation methodologies

Fast-paced startup mindset

Collaboration & communication

Job description

Reflection, a research lab, seeks to advance open and accessible intelligence. You will conduct rigorous comparative analyses, build evaluation systems, and develop frameworks that capture reasoning and alignment advances.

You will collaborate with pre-training, post-training, and applied teams to translate insights into model improvements. The role emphasizes high agency in a fast-paced startup environment, with the chance to push measurable boundaries and contribute to open foundational