Manager, AI & LLM Data Science

LexisNexis Risk Solutions; San Francisco, Northern (CA, KY); Hybrid; USD 115,000 - 192,000; Full time

Benefits offered by this job

Wellbeing initiatives

Parental leave

Study assistance

Sabbaticals

Job summary

LexisNexis Legal & Professional is seeking a Manager Data Science in San Francisco to lead a team focused on LLM training and fine-tuning, collaborating with product, engineering, and domain experts to drive business-ready AI solutions.

The role emphasizes mentorship, reproducible workflows, and innovation in model development, evaluation, and deployment within a fast-paced setting. A strong emphasis on responsible AI practices and performance optimization is expected.

Qualifications

• Strong understanding of transformer architectures and language modeling objectives.
• Experience with supervised fine-tuning (SFT) and loss masking.
• Hands-on experience with PEFT methods like LoRA/QLoRA.
• Ability to build instruction-response datasets and manage data leakage risk.

Responsibilities

• Lead the design and execution of LLM training and fine-tuning projects.
• Oversee creation of high-quality training datasets with cleaning, deduplication, annotation, and validation.
• Develop efficient fine-tuning workflows and evaluation frameworks.
• Mentor data scientists and ensure reproducible development practices.
• Collaborate with product, engineering, and domain experts to deploy and monitor models.

Skills

LLM fundamentals

Python

PyTorch

Hugging Face Transformers

SFT

PEFT

Data engineering

GPU training

Experiment tracking

Reproducibility

Tools

DeepSpeed

PyTorch FSDP

Job description

LexisNexis Legal & Professional is seeking a Manager Data Science in San Francisco to lead a team focused on LLM training and fine-tuning, collaborating with product, engineering, and domain experts to drive business-ready AI solutions.

The role emphasizes mentorship, reproducible workflows, and innovation in model development, evaluation, and deployment within a fast-paced setting. A strong emphasis on responsible AI practices and performance optimization is expected.