Data Scientist II, AI Trust

Spring Health

full-time

Posted on: 9/16/2025

Origin: • 🇺🇸 United States • Utah

✨ AI Apply

💰 $120,000 - $150,250 per year

JuniorMid-Level

PythonSaltStackSpring

About the role

Own and evolve the evaluation frameworks for our AI and ML models, translating high-level trust principles into specific, measurable tests.
Define and conduct rigorous experiments to resolve ambiguous questions about the safety, reliability, and impact of our models.
Own and conduct critical analyses and experiments to measure the real-world impact of our AI, helping to define and validate that our systems are technically robust, trustworthy, and outcome-driven.
Collaborate with engineering partners to design and build production-quality code, creating automated, scalable, and pragmatic testing frameworks based on modern best practices.
Partner with product, legal, and infrastructure teams to implement and monitor standards for trustworthy AI.
Proactively identify gaps and develop novel evaluation approaches, which may include creating synthetic test data from user traces or building lightweight processes for non-technical partners to iterate on test sets.
Synthesize complex evaluation results and industry trends into actionable insights and clearly communicate findings to diverse technical and non-technical stakeholders.

2-3 + years of relevant industry experience in data science, machine learning, or a related field.
Proficiency in Python and a solid understanding of core statistical concepts.
Proven ability to write and review production-quality code.
Proven experience in evaluating machine learning models; exposure to large language models (LLMs) is a strong plus.
Hands-on experience in one or more of the following: analyzing A/B tests or other experiments with statistical rigor; using evaluation tools (e.g., LangSmith, open-source libraries) to iteratively measure and improve model performance; building data pipelines or tools to enable collaboration on test sets.
Applied knowledge of concepts in AI ethics, such as fairness, bias, and interpretability.
A strong interest in applying data science to complex, high-stakes domains like mental healthcare.
A pragmatic and proactive approach to problem-solving, with a history of developing creative solutions to complex problems.
Exceptional communication and teamwork skills, with a proven ability to collaborate effectively with diverse, cross-functional teams.
An avid learning mindset and a passion for staying at the forefront of trends in AI safety, evaluation, and reliability.
Must be based in the Salt Lake City metro area and be willing to commute 2-3 days a week when this role transitions to a hybrid schedule in 2026.