- Work with business areas (e.g., clinical services, pharmacy, sales, marketing, product development, actuarial, operations, etc.) to understand business needs that can be addressed by extracting actionable information from available data assets.
- Build classification, regression, and potentially NLP models that are responsive to business needs
- Manage end-to-end process from understanding the business problem, selecting training data, model development and evaluation, to delivering/deploying models in the business setting.
- Present model results through summaries and presentations that tell a succinct, easy to understand story, focusing on key insights/findings relevant to stakeholders' goals and business needs.
- Contribute to ongoing development of a robust data science pipeline within the Google Cloud environment.
- Keep current with new technologies, methodologies, and applications related to ML model development and architectures.
- Contribute to manuscript preparation for peer-reviewed publications.
Qualifications and Skills
- PhD in relevant field, or Master’s degree with 2+ years of relevant experience
- Demonstrated expertise in building classification and regression ML models
- Experience with working with health care administrative claims data (ICD-10, MS-DRG, CPT/HCPCS)
- Knowledge and experience with Github, Bitbucket, or equivalent versioning systems
- Excellent written and oral presentation skills
- Experience with OHDSI tools and the OMOP data models a plus
- Record of publishing in peer-reviewed journals is a plus
Technical Skills
- Experience developing ML models and pipelines using cloud services from Google Cloud (preferred), AWS, or Azure
- Experience with ML frameworks/tools such as scikit-learn, pandas, Numpy, caret, TensorFlow, Keras, Spark, Spacy, Pytorch, XGBoost, Caffe/Caffe2, CNTK, etc.
- Experience with NLP tools/methods including OpenNLP, Stanford NLP, LSA, LDA, Gensim, FastText, NLTK, spaCy, etc.
- Knowledge of state-of-the-art ML algorithms such as BERT, ELMo, GPT, GPT-2, XLNET, T5, LSTMs, CRFs, etc., API’s, and open-source methods.
- Knowledge and experience with Github, Bitbucket, or equivalent versioning system.