Kai Nakamura
ML EngineerEmail:kai@anthropic.com
Phone:+1 415 555 0182
Location:San Francisco, CA
Professional Summary
Applied ML engineer training safety-focused LLMs. Ex-DeepMind. NeurIPS reviewer. Known for clear communication, ownership, and shipping measurable outcomes. Combines strong data fundamentals with cross-functional collaboration and a bias for execution.Experience
Jun 2023Present
Member of Technical Staff
Anthropic · San Francisco
• Trained Claude fine-tuning pipeline on 8k H100s• Reduced training cost 27% via better data curation
• Co-authored 3 constitutional AI papers
• Documented analysis playbooks used across the org
• Built models/pipelines at Anthropic impacting $3M+ decisions annually
Jan 2019Dec 2020
Data Analyst
Signal AI · London, UK
• Built models/pipelines at Signal AI impacting $2M+ decisions annually• Reduced reporting latency from days to hours for 12 stakeholder teams
• Designed experiments lifting primary metric 16%
• Owned metrics layer covering 28 KPIs with clear ownership
• Improved data quality; critical incidents down 38%
Skills
PyTorch, JAX, CUDA, Triton, Ray, Distributed Training, RLHF, DPO, Evaluation, Constitutional AIProjects
Experimentation Platform (Anthropic)
Python, SQL, dbt, Spark, Tableau, Airflow
• Documented learnings reused by peer teams
• Recognized in team/org review for execution quality
Forecasting Model
Python, SQL, dbt, Spark, Tableau, Airflow
• Documented learnings reused by peer teams
• Recognized in team/org review for execution quality
Feature Store MVP
Python, SQL, dbt, Spark, Tableau, Airflow
• Documented learnings reused by peer teams
• Recognized in team/org review for execution quality
Education
PhD, ML
MIT
Cambridge
NeurIPS Best Paper 2022Certifications
Google Data Analytics
dbt Fundamentals