LLM Engineer

Children's Hospital of Philadelphia•Philadelphia, PA

1d•$89,840 - $114,550•Onsite

About The Position

This role is designed for a technically strong, implementation-oriented data scientist who will help build and deploy large language model (LLM) and related AI solutions for real clinical and operational use cases in pathology and laboratory medicine. This position will focus on building and evaluating safe, scalable AI tools for healthcare environments, including workflow design, document understanding, model evaluation, retrieval-augmented generation (RAG), prompt engineering within larger system architectures, and reinforcement learning or iterative feedback-driven refinement. Representative projects may include laboratory question-answering tools, methods to improve the clarity and standardization of laboratory test names and reduce confusion in clinical ordering, tools to make complex pathology reports more understandable to patient families, and selected quality-support applications. The role will work closely with clinical faculty, laboratory leaders, informaticians, software developers, and hospital digital transformation partners so that what is built is not only useful within the department, but also governed, interoperable, and capable of broader enterprise scale. The ideal candidate will be excited by translating modern AI capabilities into robust, scalable, real-world products in healthcare. This position is likely to be most attractive to candidates who enjoy hands-on development, evaluation, and implementation of production-oriented AI systems in operational settings, including close collaboration with clinicians, informaticians, and software engineers.

Requirements

Bachelor's Degree in analytics, computer/data science, statistics, mathematics or related field
At least one (1) year of applied algorithm development, data science, applied statistics, machine learning, or mathematical modeling projects experience
Basic proficiency with Retrieval-Augmented Generation (RAG) and Reinforcement Learning (RL)
Basic proficiency with Cloud machine learning platforms, such as Azure AI Studio / Azure OpenAI / Microsoft Foundry, AWS Bedrock, Google Cloud Vertex AI, OpenAI API, Anthropic, Databricks, Snowflake Cortex, Hugging Face, etc.
Basic proficiency with developing and implementing AI agents
Basic proficiency with validation, monitoring, safety (identifying and addressing vulnerabilities, biases, and edge cases before deployment)
Fundamental knowledge of orchestration frameworks and/or multi-agentic work
Basic proficiency with formulating analysis plans, selecting appropriate methods, and implementing data analysis and machine learning pipelines
Basic proficiency with programing languages (Python, R, Java, TypeScript, C/C++, C#, etc)
Basic proficiency with writing code in applied academic or professional projects
Basic proficiency with creating informative visualizations for complex, high dimensional data
Basic proficiency with machine learning models, both classical machine learning (generalized linear models, non-linear models, tree-based methods, etc) as well as deep learning
Fundamental knowledge of relational databases (Postgres, MySQL)
Fundamental knowledge of web services application programming interfaces
Strong verbal and written communications skills
Ability to maintain confidentiality and professionalism
Ability to work independently with minimal supervision
Ability to gather, analyze and make recommendations/decisions based on data / best practices

Responsibilities

Implement computational algorithms and experiments for test and evaluation; interprets data to assess algorithm performance.
Develop high-quality code implementing models and algorithms as application programming interfaces or other service-oriented software implementations.
Participate in communication of research methods, implementation, and results to varied audience of clinicians, scientists, analysts, and programmers.
Work closely with applications research group to translate models and algorithms into engineered production applications.
Contribute to manuscript writing for results publication, authors abstracts, and presents at professional conferences.