Data Scientist (AI)

Location: Bay Area, United States; hybrid

Company

hirify.global is an AI evaluation platform founded by researchers from UC Berkeley's SkyLab that measures how frontier models perform in real-world workflows.

What you will do

• Explore large, complex datasets to identify patterns, biases, causal relationships, and model-performance insights.
• Formulate hypotheses about data quality, evaluation outcomes, and model behavior, then design experiments to test them.
• Build reproducible analysis pipelines with Python, Pandas, NumPy, and Spark.
• Partner with ML researchers and engineers to develop metrics and analyses for model evaluations across domains, prompts, and tasks.
• Develop causal reasoning frameworks and statistical methods that explain model behavior and improve evaluation reliability and interpretability.
• Communicate findings to technical and non-technical partners through clear analyses and written content.

Requirements

• 6+ years of experience in data science, ML analytics, or applied research, preferably in AI, ML, or large-scale data environments.
• Strong Python skills with deep experience in Pandas, NumPy, and distributed frameworks such as Spark.
• Expertise in statistical modeling, causal inference, and experimental design.
• Experience analyzing data distributions, sample quality, and distribution shifts.
• Strong communication skills and close collaboration with ML researchers and engineers.
• Senior level or higher is required for this role.

Nice to have

• Background in AI model evaluation.
• Experience with LLM outputs, LLM-as-a-judge, embeddings, or other large-scale model artifacts.
• Experience with A/B testing.

Culture & Benefits

• Competitive compensation and equity aligned with the candidate's permanent work location.
• Comprehensive medical, dental, and vision coverage with additional wellness support.
• Opportunity to work on cutting-edge AI with a small, mission-driven team.
• Culture focused on transparency, trust, craftsmanship, curiosity, and community impact.