Remote Lead Data Scientist — AI/ML & NLP (LLMs)

S&P Global Inc.; New York (NY); Remote; USD 190,000 - 206,000; Full time

Job summary

Standard & Poor's Financial Services LLC is seeking a Lead Data Scientist to guide the design and deployment of ML, NLP, and LLM-based solutions across batch and streaming pipelines. You will architect data ingestion, preprocessing, retrieval, RAG, and prompt engineering using PyTorch, Hugging Face, and LangChain.

Collaborate with data science, MLOps, and technology teams to ensure robust evaluation, scalable production deployment on AWS, GCP, or Azure, and governance with strong documentation.

Qualifications

• Master’s degree in Computer Science, Data Science, or a related IT field.
• 3 years of experience in Python programming and ML frameworks.
• Experience with PyTorch, Hugging Face Transformers, and LangChain.
• Experience with LLM architectures and training/fine-tuning for domain-specific apps.
• Experience with NLP solutions and information retrieval using vector embeddings and semantic search.

Responsibilities

• Lead design and development of ML, NLP, and LLM-based pipelines.
• Architect data ingestion, preprocessing, search/retrieval, RAG, fine-tuning, and prompt engineering.
• Collaborate with data science, MLOps, and engineering teams to meet requirements.
• Deploy models to production with high reliability, scalability, and performance on cloud platforms.
• Develop and maintain documentation, governance, and reproducibility across teams.

Skills

Python programming

ML frameworks

NLP

LLMs

Transformer models

Model deployment

MLOps

stakeholder management

Education

Master's degree in Computer Science, Data Science, or related IT field

Tools

PyTorch

Hugging Face Transformers

LangChain

AWS

GCP

Azure

Job description

Standard & Poor's Financial Services LLC is seeking a Lead Data Scientist to guide the design and deployment of ML, NLP, and LLM-based solutions across batch and streaming pipelines. You will architect data ingestion, preprocessing, retrieval, RAG, and prompt engineering using PyTorch, Hugging Face, and LangChain.

Collaborate with data science, MLOps, and technology teams to ensure robust evaluation, scalable production deployment on AWS, GCP, or Azure, and governance with strong documentation.