About the Role:

Grade Level (for internal use):

The Role: Sr Data Scientist– NLP, LLM and GenAI
 

The Team:

The Capital IQ Solutions Data Science team supports the S&P Capital IQ Pro platform with innovative Data Science and Machine Learning solutions, utilizing the most advanced NLP Generative AI models. This role presents a unique opportunity for hands-on ML/NLP/Gen AI/LLM scientists and engineers to advance to the next step in their career journey and apply their technical expertise in NLP, deep learning, Gen AI, and LLMs to drive business value for multiple stakeholders while conducting cutting-edge applied research in LLMs, Gen AI, and related areas.

Responsibilities: 

• Design solutions utilizing NLP models including chat assistants and RAG systems.
• Design and develop custom NLP LLM Models including both prompt engineering techniques and model fine-tunning and alignment (SFT, RLHF, DPO)
• NLP Model evaluation using both human-supported and synthetic evaluation methods and metrics.
• Deploy NLP models ensuring latency, reliability, and scalability.
• Discover new methods for prompt engineering, model fine-tuning, quantization and latency optimization, document embeddings and chunking.
• Collaborate closely with product teams, business stakeholders, and engineers to ensure smooth integration of NLP models into production systems.
• Troubleshoot complex issues related to machine learning model development and data pipelines and develop innovative solutions.
• Actively research, explore and identify the latest relevant methods and technologies

Essential Skills: 

• Degree in Computer Science, Mathematics or Statistics, Computational linguistics, Engineering, or a related field. 
• Good understanding of machine learning and deep learning methods and their mathematical foundations
• 1+ years of professional experience in Advanced Analytics / Data Science / Machine Learning
• 1+ years hands-on experience developing NLP models, ideally with transformer architectures. 
• Demonstrated experience with Python, PyTorch, Hugging Face or similar tools. 
• Mastery of Python and ability to write robust and high standard, testable code
• Knowledge of developing or tuning LLMS

Desirable Experiences:

• 1+ years of experience with implementing information retrieval systems.
• Experience with contributing to Open Source initiatives or in research projects and/or participation in Kaggle competitions.
• Publications related to Machine Learning or Deep Learning

Person Sills

• Ability to work in a team
• Able to report progress and summarize issues to a less technical audience
• Curious and open-minded attitude to new approaches

About S&P Global Market Intelligence

At S&P Global Market Intelligence, a division of S&P Global we understand the importance of accurate, deep and insightful information. Our team of experts delivers unrivaled insights and leading data and technology solutions, partnering with customers to expand their perspective, operate with confidence, and make decisions with conviction.

For more information, visit www.spglobal.com/marketintelligence.

What’s In It For You?

Our Purpose:

Progress is not a self-starter. It requires a catalyst to be set in motion. Information, imagination, people, technology–the right combination can unlock possibility and change the world.

Our world is in transition and getting more complex by the day. We push past expected observations and seek out new levels of understanding so that we can help companies, governments and individuals make an impact on tomorrow. At S&P Global we transform data into Essential Intelligence®, pinpointing risks and opening possibilities. We Accelerate Progress.

Our People:

We're more than 35,000 strong worldwide—so we're able to understand nuances while having a broad perspective. Our team is driven by curiosity and a shared belief that Essential Intelligence can help build a more prosperous future for us all.

From finding new ways to measure sustainability to analyzing energy transition across the supply chain to building workflow solutions that make it easy to tap i