Remote, 12-month contract with potential to renew
As a real-world data machine learning analyst, this individual will conduct hands-on programming using real-world data (RWD) to deliver results for machine learning and RWD-related queries and studies, under the guidance of the Director/Senior Manager of Health Economics and Outcomes Research (HEOR) ML scientists.
Responsibilities & Deliverables:
- Understand machine learning analytic requests (e.g., predictive models, clustering analysis) specified by HEOR ML scientists, and provide deliverables in the form of Python/R programming codes, Jupyter Notebooks, R Markdown files, emails, Excel tables, Word documents, or PowerPoint presentations within pre-specified timelines.
- Proactively clarify requests and ensure accurate implementation of protocols and ML model specifications from the requestors.
- Maintain detailed documentation and generate results with clear and professional presentation.
- Ensure high-quality work products with the ability to effectively communicate methods and findings.
- Ability to work independently and efficiently to meet aggressive timelines when needed.
- Adapt to rapidly changing priorities.
Skills:
- Expert-level proficiency in Python and/or R programming languages
- Expert knowledge of Python machine learning packages (e.g., scikit-learn, XGBoost, PyTorch)
- Expert skills in data preprocessing and feature engineering approaches for preparing machine learning-ready dataframes
- Proficiency with the Databricks platform
- Proficiency with version control systems (e.g., GitHub, BitBucket)
- Familiarity with data science reporting tools such as Jupyter Notebook and R Markdown
Qualifications:
- A Master’s degree, or higher, in computer science, data science, informatics, statistics/biostatistics, or a quantitative area with a strong focus on the application of data science to healthcare data with 2+ years of hands-on programming experience
- Must have hands-on RWD/RWE and ML experience with data sources such as EHR (e.g. concertAI, Flatiron) and large-scale administrative claims data (e.g., MarketScan, Optum, Medicare LDS SAF, Komodo claims)