Andrew Johnson

Experience

• 

  University of Wisconsin-Eau Claire

  Jan 2006 - May 2010

  Undergraduate Research Assistant

  Worked with multiple research projects to initiate and develop computational components to in-flight experimental and observational analyses. Research projects were involved in both spectroscopic chemistry and astrophysical stellar winds.• Supplemented two ongoing experimental research programswith computational and statistical components to validate experimental data• Generated simulations to produce data used towards the identifi cationof molecular frequency states• Identified previously unknown high-energy frequency modesin carbon-based molecules• Tested the robustness of the mathematical routines in the Q-Chem softwarefor computational chemistry, and collaborated with developers to assistwith software enhancements• Repurposed simulation software intended for nuclear fusion forastrophysical simulations of stellar plasmas• Used applied mathematics and statistics from simulated data to enhanceunderstanding of scientifi c parameters in stellar winds Show less

• 

  University of Minnesota-Twin Cities

  Jun 2010 - Jun 2014

  Graduate Research Assistant

  Involved in graduate research projects applied to magnetohydrodynamic simulations applied to intracluster medium analyses.• Wrote an extensive, parallelized, statistical suite intended to analyzein-house generated terabyte scale datasets• Generated mathematical models in conjunction with statistical analysesfor describing spatial distributions of observational galaxy cluster data• Used Fourier analysis to categorize spatial scales in computationallygenerated uid dynamic simulations through applied Kolmogorov statistics• Enhanced community understanding of several key parameters in thecurrently accepted paradigm for galaxy cluster physics• Gave ~10 major scienti c presentations to highly technical audiences Show less

• 

  Tightrope Media Systems

  Oct 2014 - Feb 2016

  Digital Signage Support

  Principal support specialist for the Carousel line of products.• Used sales data to facilitate improvements in capturing revenue fromsoftware renewals of existing customer base• Created geospatial analytics by utilizing transactional requestsgenerated by customer data streamed through cloud API• Formulated a timing model for efficient SQL transactions in core product• Designed updated ticketing systems for improving support workow• Used incident ticket data to design pricing models for proposed support costs• Responsible for ~300 support tickets per month utilizing e-mail, phone,and web conferencing Show less

• 

  Optum

  Jan 2016 - now

  Lead Data Scientist and AI/ML contributor on Care Management platform modernization initiative.• Developed models (xgboost) used in predicting member propensityto enroll for both disease and case management programs. Models deployedto production utilizing online data for real-time scoring based off high-riskhealth event triggers• Developed recommendation model (DQN), using personalized data profiles formembers and assigning affinities to developed nurse personas. Production Modelresulted in 15% improvement of conversion for targeted members intohigh-risk disease management programs• Created a POC cognitive agent using fine-tuning of pretrained BERT modelsand knowledge graph for providing real-time monitoring of nurse-to-member chatinteractions and highlighting key topics from Omaha Taxonomy. Presentedwork at 2020 Graph + AI World Conference• Contributed to the design of AI/ML model frameworks for real timeupdate of feature stores, and real-time execution of model inference• Worked consistently with Sr. Leaders to help provide AI/ML directionof blue chip Care Management functionality. Provided training sessions,seminars, and assisted in acquisition of talent for AI/ML initiatives Show less Full-stack data scientist and principle designer of solutions utilizing machine learning, analytics, data aggregation, and data warehousing for continuous improvement and automation around operational workflows.• Designed, developed, and implemented a machine learning algorithm to automateremediation of ~100000/year failed file-based eligibility transactions• Lead an analytical initiative to identify cost improvements for first callresponse support cases, leading to the identification and proposal ofreduction targets for $1m/year savings opportunities• Designed, developed, and implemented a data aggregation program anddata store for use with multiple disparate data sources containing incident,eligibility, Salesforce, biometric, and activity related data• Principal analyst for a rigorous investigation providing recommendationstowards enhanced user experience and increased user impact forwellness incentive programs• Used natural language processing to do a rigorous analysis on Salesforcecase intake data used to identify cases at high risk for escalation Show less Manage and coordinate activities and process improvements around the software development life cycle for core portal applications.• Developed computational and analytical tools to provide reportingcapabilities and increase the efficiency of development project hour estimation• Consolidated delivery related data across multiple sources and coordinatedwith organizational improvements of SDLC processes for over 100 applications• Directed and executed the migrations of over 500 application repositories fromSVN to Github enterprise• Coordinated the successful go-live for a Medicaid management portal givena 2 month SLA on delivery• Principle manager for remediation of over 5000 pre-existing securityvulnerabilities in a legacy web application• Principle project manager for the integration of CICD, cloud infrastructuremigration, and SDLC automation on a large-scale web application Show less

  • Prin. Data Scientist, Technical Leadership Career Path

    Jun 2020 - now

  • Sr. Data Scientist

    Jun 2018 - Jun 2020

  • Technical Delivery Manager

    Jan 2016 - Jun 2018