Senior AI/ML Data Scientist, Natural Language Processing

Job Description

Position Description:

Senior Artificial Intelligence and Machine Learning (AI/ML) Data Scientist, Natural Language Processing

The Senior AI/ML Data Scientist – Natural Language Processing (NLP) role involves helping to develop and deploy production-grade NLP products for unstructured and semi-structured data from across our company’s research and development pipeline. These models and workflows will help solve real-world problems and contribute to Artificial Intelligence and Machine Learning (AI/ML) in therapeutic research and development.

Key focus areas will include the scalable deployment of ML and Generative AI approaches (such as Large Language Models, or LLMs) for surfacing insights from proprietary unstructured research data and biomedical literature, as well as developing fit-for-purpose approaches for the likes of text classification, relation extraction, and entity linking.

Additional job details:  

The types of datasets we focus on are both internal (e.g., electronic lab notebooks, safety reports, regulatory documents, clinical results) and external (e.g., public literature and Electronic Medical Records). In addition to new tool development, we often consult with some of our 5,000+ stakeholders (scientists, engineers, regulatory liaisons, data scientists, etc.) on their own projects, as well as additional stakeholders from across our multi-national company. We strive to enhance data science, NLP, and AI literacy across these groups. As part of our work, we have opportunities to coauthor presentations, reports, manuscripts, and/or public code releases. 

The position is embedded in a cross-disciplinary team of data scientists, bioinformaticians, and engineers that are all focused on using cutting-edge software, AI/ML, and data science techniques to drive drug discovery and development. 

You enjoy:  

• Building novel tools that enable the discovery, development, and delivery of new therapeutics to patients in need 
• Understanding real-world challenges and developing automated data solutions for them 
• Opportunities to directly interact with users of your data science, ML, and AI products 
• Evaluating, developing, testing, and deploying new techniques for natural language understanding 
• Freedom to propose projects that interest you and to collaborate cross-functionally on delivery 
• Staying updated on the newest methods in NLP, ML, and generative AI 
• Sharing the approaches you implement and their impact with internal company audiences and externally 

Position Qualifications:

Education Minimum Requirement:

• A PhD in data science, AI/ML/LLM engineering, computer science, semantic engineering or a related discipline OR
• M.S. with 2 years industry experience OR
• B.S. with 5 years industry experience focused on NLP, data science, AI/ML/LLM engineering, computer science, semantic engineering or a related discipline 

Required Experience and Skills:

• 2 years experience with Python, Spark, or related frameworks in AI, machine learning, data science, data engineering, or similar context 
• 1 years experience with Natural Language Processing, Generative AI, or related techniques for machine understanding of natural language (i.e., written text, omics data, or similar).  

Preferred Experience and Skills:

• Fluency in Python programming, version control and collaboration with git, environment management (e.g., poetry, conda, docker), standard Python packages (e.g., pandas, numpy, matplotlib), and at least one ML framework (e.g., pytorch, tensorflow, fairseq) 
• Experience with scalable data engineering frameworks such as Apache Spark and orchestration frameworks such as Airflow, and/or experience with semantic search and retrieval frameworks (e.g., development and benchmarking of embedding models and retrieval approaches in the context of Retrieval Augmented Generation, RAG). 
• Experience with ML model deployment and operations (e.g., DevOps, MLOps, LLMOps), including CI/CD workflows and tooling (e.g., Github actions) 
• Experience with standard operations on non-relational (e.g., Elasticsearch/Opensearch, MongoDB, Neptune), relational databases (e.g., PostgreSQL), and vector databases (e.g., pgvector, Elasticsearch dense vectors) and deployment of APIs and web applications (e.g., flask, fas