Senior Language Modeling Research Engineer - Artificial Intelligence

Bloomberg - New York, NY

Full-time - Senior

New York, NY

251-500 employees

Broadcasting and Content Providers

About the position

The Senior Language Modeling Research Engineer will be part of Bloomberg's Engineering AI department, focusing on the development and application of Large Language Models (LLMs) to enhance AI-driven customer-facing products. This role involves collaborating with Machine Learning and Software Engineers to innovate and improve financial NLP models, ensuring high accuracy and low latency in processing vast amounts of structured and unstructured information.

Responsibilities

• Collaborate with colleagues on building and applying LLMs for production systems and applications
• Write, test, and maintain production quality code
• Train, tune, evaluate and continuously improve LLMs using large amounts of high-quality data
• Demonstrate technical leadership by owning cross-team projects
• Stay current with the latest research in AI, NLP and LLMs and incorporate new findings into models and methodologies
• Represent Bloomberg at scientific and industry conferences and in open-source communities
• Publish product and research findings in documentation, whitepapers or publications to leading academic venues

Requirements

• Practical experience with Natural Language Processing problems
• Familiarity with Machine Learning, Deep Learning and Statistical Modeling techniques
• Ph.D. in ML, NLP or a relevant field or MSc in CS, ML, Math, Statistics, Engineering, or related fields and 2+ years of relevant work experience
• Experience with Large Language Model training and fine-tuning frameworks such as PyTorch, Huggingface or Deepspeed
• Proficiency in software engineering
• Understanding of Computer Science fundamentals such as data structures and algorithms
• Excellent communication skills and the ability to collaborate with engineering peers and non-engineering stakeholders
• A track record of authoring publications in top conferences and journals is a strong plus