LaunchPoint   •  

Nashville, TN

Less than 5 years

Posted 241 days ago

This job is no longer available.

Job Description:

About Us: Discovery Health Partners offers payment and revenue integrity solutions that help health payers improve revenue, avoid costs, and enhance the member experience. We offer a unique combination of deep healthcare expertise and analytics-powered technology solutions to help our clients improve operational efficiency, achieve financial integrity, and generate measurable results.

About You: Do you love all things Data Science, start to finish:

  • Wrangling and repurposing data at 1-100TB scale;
  • Rendering data suitable for modeling and learning;
  • Engineering innovative features;
  • Picking the right methods and using them to find signal in what to human eyes is overwhelming complexity or even noise;
  • Validating that signal is being robustly, replicably pulled out;
  • Transforming your models into deployable artifacts;
  • Helping shepherd your handiwork into production;
  • Monitoring model performance and adjust models if and as needed; and
  • Keeping all these skills at the cutting edge of technology?

Do you thrive in the pace and flexibility of a startup environment, where the leading edge just starts to turn into the bleeding edge? Do you thrive on cross-functional conversation in which you’re one of several contributors? Have you started developing the engineering acumen to have a solid skill base in which you’re confident, and yet also a growing appreciation that you can never master it all? Do you have strong respect for and clear ideas about what data and data-exploitation methods can and cannot do, and where the pitfalls are? If you think these are some of the right questions, and your qualifications stack up well relative to the following, then please reach out to us.

Key Role/Responsibilities:

To more broadly support and grow our base of healthcare payer customers, Discovery Health Partners is creating an internal startup to develop an entirely new line of products that deliver actionable insights yielding strong customer ROI toward improving their payment integrity and revenue optimization. For this effort, the Data Scientist develops models and algorithms, using data-aggregating, statistical, machine-learning and advanced mathematical techniques; and delivers those as plug-and-play deployable artifacts that yield the value-creating, actionable insights driving our new line of analytic-oriented products. In this vital role, you will master some of the complex business problems our clients, our healthcare system, as well as our own data challenges, deploying and growing your mastery in developing solutions for these problems. Specifically, the Data Scientist:

  • Designs, develops, implements, and maintains value-add analytics for Discovery Health Partners’s product portfolio, primarily by leveraging large healthcare data-sets, to produce valid, replicable, and defensible analytic content.
  • Monitors and tunes data models continuously to ensure they remain accurate as additional data is analyzed and as our products change healthcare-related behavior that we’re modeling.
  • Stays current with and researches, evaluates, recommends, and utilizes appropriate data mining, predictive modeling, and/or statistical analysis tools and methodologies to perform relevant analyses.
  • Performs ad hoc requests.
  • Adheres to and supports enterprise standards for the design and implementation of analytics.
  • Engages in quarterly back-sweeps to identify and clean-up the inevitable technical debt that comes with rapid growth and yet also eventually impedes it.
  • Maintains effective, proactive, energetic working relationships within team and among internal customers.
  • Engagingly communicates results and technical considerations to non-technical stakeholders.

Required Skills

Strong ability in wrangling data and features, developing and validating models, and delivering deployable artifacts, at 1-100TB scale, and using open-source tooling.

  • Strong skills developing and validating performant models in Python, R and/or Scala.
  • Concomitantly strong programming skills writing performant python, including to run models in other languages from within python.
  • Robust mathematical skill, with strong working knowledge in some and the ability quickly to be competent in the rest of the gamut of modeling techniques and associated tooling, for examples:
    • Established statistical methods for data reduction (e.g., factor analysis, PCA) and for prediction (e.g., regression);
    • Established machine learning methods for feature extraction (e.g., neural nets), for data reduction (e.g., clustering) and for prediction (e.g., random forest, graph/network analysis);
    • Cutting edge methods like nonlinear NLP techniques (e.g., Poincare balls in hyperbolic spaces / Riemannian geometries), complex time-series (e.g., GARCH), and topological analysis of data shape (e.g., persistent homotopy, Morse-Smale complexes).
  • Strong skills writing performant SQL. Redshift and Postgres-with-MPP both plusses.
  • Strong skill producing visualizations of data and results.
  • Strong shell scripting skill a strong plus.
  • Experience in a hybrid Linux/Windows environment a plus.
  • Experience in robustly secure environments a plus. Experience with healthcare data a strong plus. Experience in AWS a strong plus.
  • Strong in at least one object-oriented language. Python and Scala are big plusses, even more so with experience in functional programming.
  • Technical plusses include facility with: event-driven service-oriented solutions, Agile methodology, code that automates coding, DFAs, DSL/parser-generators.
  • Solid skills communicating technical considerations to non-technical stakeholders. Strong sense of humor a plus.
  • Strong skill in and drive for collaborative teamwork across diverse disciplines and backgrounds, working through needed skepticisms and strong opinions to get to a prudent and aggressive Yes.

Required Experience

  • A minimum of 3 years of experience in industry, wrangling data and features, developing and validating models, and delivering deployable artifacts, at 1-100TB scale, using open-source tooling.
    • Masters required, Ph.D. preferred, in machine learning, statistics, applied mathematics or strongly quantitative, stochastic discipline.
    • Additional clinical degree (M.D., M.R.N. or similar.) a very strong plus; or experience with biomedical research a plus.