AI Scientist Intern at Regence… · Portland

Summary: Regence BlueCross BlueShield of Oregon is seeking an AI Scientist Intern for a 12-week full-time internship starting in May or June 2026. The intern will work under the mentorship of an AI Scientist, focusing on designing, developing, and implementing data-driven models and algorithms to solve business problems in the healthcare domain. Responsibilities: - Researches, designs, develops, and implements data-driven models and algorithms using machine learning, deep learning, statistical, and other mathematical modeling techniques - Trains and tests models and develops algorithms to solve business problems - Adheres to standard best-practices and establishes principled experimental frameworks for developing data-driven models - Develops models and performs experiments and analyses that are replicable by others - Uses open-source packages when appropriate to facilitate model development - Identifies, measures, analyzes, and visualizes drivers to explain model performance (e.g., feature importance, interpretability, bias and error analysis), both offline (in the development phase) and online (in production) - Uses appropriate metrics and quantified outcomes to drive model and algorithm improvements - Analyzes, diagnoses, and resolves bugs in production machine learning models and systems - Evaluates model/use case feasibility by quickly generating prototypes - Takes models from prototype stage and improves performance as needed - Writes clean, well-commented, tested, version-controlled, and maintainable python code - Collaborates with team members and Cambia business partners - Actively participates in group meetings and discussions - Communicates effectively both orally and in writing with both technical and non-technical audiences - Keeps current with the state of the art in machine learning and AI and its application to healthcare - Keeps current with evolving commercial and open-source tools, techniques, and brings these practices to projects - Over time develops familiarity and insight with various subdomains of healthcare data Required Qualifications: - Demonstrated knowledge of data science, machine learning, and modeling - Ability to use well-understood techniques and existing patterns to build, analyze, deploy, and maintain models - Effective in time and task management - Able to develop productive working relationships with colleagues and business partners - Strong interest in the healthcare industry - Ability to read, understand, and apply the latest research to enhance our products where possible - Strong mathematical foundation and theoretical grasp of the concepts underlying machine learning, optimization, etc - Demonstrated understanding of how to structure simple machine learning pipelines (e.g., has prepared datasets, trained and tested models end-to-end) - Classic ML algorithms (e.g., linear and logistic regression, decision and boosted trees, SVM, collaborative filtering, ranking) - Approaches (e.g., supervised, semi-supervised, unsupervised, reinforcement learning, regression, classification, time series modeling, transfer learning) - Foundational ML concepts such as objective functions, regularization and overfitting - Data partitions (train/dev/test) and model development - Hyperparameter tuning and grid search - Evaluation concepts (metrics, feature importance, etc.) - Familiarity with standard python packages (scikit-learn, XGBoost, TensorFlow, PyTorch, etc.) - Familiarity with structure of machine learning pipelines - Experience with natural language processing (NLP) techniques such as tokenization, part-of-speech tagging, named entity recognition, sentiment analysis, and machine translation - Familiarity with pre-trained models and frameworks like BERT, GPT, and spaCy for NLP tasks - Understanding of text preprocessing, feature extraction, and embedding methods - Experience with NLP libraries and tools (NLTK, Hugging Face Transformers, etc.) - Capable of building and refining NLP models for tasks such as text classification, language generation, and information extraction - Awareness of the ethical considerations and challenges in deploying NLP models - Large Language Models (LLMs) and their capabilities (e.g, in-context learning, few-shot learning, zero-shot learning) - Prompt engineering techniques and best practices - Fine-tuning approaches (e.g, full fine-tuning, parameter-efficient methods like LoRA, QLoRA) - Retrieval-Augmented Generation (RAG) and knowledge integration - Evaluation methods for generative models (e.g, perplexity, BLEU, ROUGE, human evaluation, LLM as a Judge) - Alignment techniques (e.g, RLHF, constitutional AI, red-teaming) - Multimodal generative models (text-to-image, text-to-video, multimodal understanding) - Responsible AI considerations specific to generative models (e.g, bias, hallucinations, safety) - Strong foundation in data analysis - Research and experiment design - Visualization with data - Answering questions with data - Strong python programming skills - Familiarity with standard data science packages - Familiarity with standard software development best practices - Strong SQL skills a plus - Understanding of standard algorithms and data structures (ex. search & sort) and their analysis Required Skills: Machine Learning, Deep Learning, Statistical Modeling, Data Science, Healthcare Payer Domain Knowledge, Python Programming, SQL, Natural Language Processing, Generative AI, Data Analysis, Algorithm Design, Software Development Best Practices Benefits: Earn a competitive salary and enjoy generous benefits while doing work that changes lives, Grow your career with a company committed to helping you succeed, Give back to your community by participating in Cambia-supported outreach programs, Connect with colleagues who share similar interests and backgrounds through our employee resource groups, Work from home options for most of our roles