Data Scientist II

To be part of our organization, every employee should understand and share in the YNHHS Vision, support our Mission, and live our Values. These values - integrity, patient-centered, respect, accountability, and compassion - must guide what we do, as individuals and professionals, every day.

The data scientist applies analytic skills to understand forecasting problems facing the health system, suggests pragmatic solutions, and develops applications implementing solutions in daily operations. The job requires excellent technical skills in manipulating and analyzing large volumes of data. The scientist demonstrates competency in using a variety of mathematical, statistical, and machine learning approaches to develop predictive solutions. In addition, the scientist must display excellence in teamwork and communication in all aspects of their work. EEO/AA/Disability/Veteran.

EEO/AA/Disability/Veteran

• Critically evaluates all requests for predictive applications.
  • Critically evaluates all requests for predictive applications.
  • Drafts specification document for new projects outlining needs, methods to address those needs, development timeline, and deliverables.
  • Reviews requests and proposed development plans with JDAT leadership with an evaluation of project feasibility.
  • Drafts approach to implement project in daily operations.
  • Works with other JDAT analysts to pull data for predictive projects.
  • Cleans and validates data using exploratory data analysis and collaboration with customers and clinical subject matter experts.
  • Finds the best predictive analytics approach and algorithm(s) to address above needs.
  • Develops predictive application and a user interface for pilot testing by requestors.
  • Drafts report of the predictive application's performance metrics for leadership and requestors.
  • Works closely with requestors to operationalize application.
  • Develops final production application and drafts monitoring and maintenance plan.
  • Drafts user documentation for production application, including the above plans.
  • Takes responsibility and ownership for production applications. Works independently to assure smooth operation.Independently manages work effort among several deployed and in-production predictive applications. When concerns regarding effort bandwidth arise, escalates issues to leadership.
  • Delivers applications according to timeline agreed upon by JDAT leadership and requestors.
• Teamwork
  • Integrates with other members of the analytics team. Takes initiative to communicate, establishes and maintains relationships, and understands roles and processes.
  • Engages health system customers cordially and responds to requests in a timely manner.
  • Takes an active approach in ensuring customer satisfaction, including taking initiative for follow-up
• Documentation
  • Keeps organized documentation on all predictive projects, from initial specification and development plans to project progress and milestones and finally maintenance and monitoring plans with follow-up notes on monitoring.
  • Drafts documentation for customers on application usage. Reviews drafts with JDAT leadership before publishing final version.
  • Documents all project code well to ensure maintainability and understanding by team members.
  • Version controls all project code and documentation and ensures that central JDAT repository has an up-to-date copy
• Leadership
  • Displays strong organizational, problem solving and listening skills, attention to detail, innovative thinking and ability to inspire others.
  • Provides leadership and coordination to junior data scientists.
  • Provides training to other IT staff and user clients as appropriate.
  • Coordinates interactions and activities of vendors, both on-site and off-site.
• Continuing Education
  • Regularly searches and seeks out other data science practitioners, whether online, locally, or through conferences. Learns from their experiences, communicates new findings to team members and leadership, and attempts to use new techniques.
  • After reviewing with JDAT leadership, pursues further career related coursework, either online or through physical attendance.
    

EDUCATION

Masters in data science, mathematics, statistics, engineering or closely related field required with work relevant to data science, such as modeling, will be given extra preference. PhD preferred

EXPERIENCE

Three years of experience statistical analysis or related work with at least one year experience in health care analytics. Epic certification in Analytics highly preferred. Fluency with EHR tools used for analytics, especially predictive analytics such as example fluencies: Epic Reporting Workbench, Caboodle and Clarity data models,utilities to work with Epic Cloud, understanding of Epic provided predictive models and use cases. Experienced in programming with R to build machine learning models using a variety of techniques both linear and non-linear, for example, GLMNet, random forest and xgboost. Similar experience with Python preferred. Experience with building C and C++ routines that can connect with R to optimize numerical routines is highly preferred. Familiarity with low level graphics libraries in R to develop novel data visualization methods a plus. Ability to apply all of the above to business problems and delivering solutions is highly valued. Experienced in applying abstraction and generalization to extract overall features of a project and then communicating these features in simple, outline form to others, especially clinical or administrative staff, who are seeking to understand the implications or ramifications of a predictive analytics project. Experience in communicating the outcome and results of a project to an academic audience a plus.

SPECIAL SKILLS

• Excellent written and oral communication skills preferred
• Experience in using the command-line interface in Windows and/or Linux.
• Experience with open-source command-line utilities useful in data transformation and cleaning
• Experience in using the git version control system.
• Experience with good software engineering practices: writing small discrete procedures, unit testing, and writing libraries or packages
• Ability to format written communications using the LaTeX document preparation system a plus.

PHYSICAL DEMAND

NA