Lead Data Scientist on the Predictive Scheduler project presented at HIMMS23 (https://hde.himss.org/view_mode_page/node/153967/schedule_modal).Builds and maintains models to optimize the staffing and scheduling of various service lines across the organization’s 50 plus hospitals and 1000 plus clinics. These optimization models have the potential to save millions of dollars over previous state productivity per KPIs in ongoing user testing / pilots.Translates business needs and… Show more Lead Data Scientist on the Predictive Scheduler project presented at HIMMS23 (https://hde.himss.org/view_mode_page/node/153967/schedule_modal).Builds and maintains models to optimize the staffing and scheduling of various service lines across the organization’s 50 plus hospitals and 1000 plus clinics. These optimization models have the potential to save millions of dollars over previous state productivity per KPIs in ongoing user testing / pilots.Translates business needs and goals into mathematically optimizable models ranging from mixed integer linear programming models to constraint programming models. Leverages CPLEX and Google OR-Tools for optimization.Builds and maintains models that forecast KPIs and other important metrics such as the average daily census, hourly patient volumes, hourly boarder patient volumes, and more with high accuracy (80% or greater). Leverages Meta’s Prophet model, XGBoost, etc. in forecasting work.Deploys and maintains instances of Microsoft Azure’s OpenAI, Cognitive Search, and Form Recognizer services to create privately hosted, world-class chat bots on top of internal enterprise data.Employs Hugging Face pretrained models and transfer learning to conduct a variety of NLP tasks including sentiment analysis, sarcasm detection, and more on top of enterprise survey results.Hosts and attends cross-functional meetings with stakeholders (business leaders, nurse managers, schedulers, service engineering & security). Show less

Leveraged SQL, Python, and Azure to build and implement machine learning models in the tax and audit practices. Extensively employed NLP methods (word embeddings, topic modeling, etc.) and neural networks in my modeling.Developed and productized models enabling the automated classification of general ledger items into tax groups, subgroups, and M-adjustments. This cut the time for tax accounts to classify general ledgers by upwards of 75% per KPIs.Developed and productized… Show more Leveraged SQL, Python, and Azure to build and implement machine learning models in the tax and audit practices. Extensively employed NLP methods (word embeddings, topic modeling, etc.) and neural networks in my modeling.Developed and productized models enabling the automated classification of general ledger items into tax groups, subgroups, and M-adjustments. This cut the time for tax accounts to classify general ledgers by upwards of 75% per KPIs.Developed and productized models for the prediction of charge hours for proposed client engagements. This allowed the assurance team to optimize the budgeting of their time and which projects they took on.Continuously met with stakeholders throughout the development lifecycle to understand and meet their needs.Leveraged Python’s Dash framework and Alteryx to build frontends to conduct strenuous user acceptance testing of models. Iteratively developed based on user feedback. Scaled user-accepted models by deploying to a production Azure Kubernetes Service and helped oversee its maintenance.Monitored production models and ran diagnostic reports using Azure Logic Apps. Established mechanisms to detect data drift. Show less

Worked extensively with the Chief Information Officer, Director of IT Strategy, and Director of Sales to build an application which recommends cross-selling opportunities. Received praise and thanks upon presenting both the application and findings regarding data quality to the CEO and executive team.Engaged in the creation of an end-to-end machine learning pipeline employing Azure Machine Learning and Azure Kubernetes Service.

Leveraged SQL to organize over a million client data entries across thousands of fields into a structure consumable for machine learning.Brainstormed and addressed analytical scenarios using CRM, ERP and benchmark data in Power BI.

Applied leadership and organizational skills routinely in presiding over the opening shift.Signed off and stocked all deliveries.Registered and prepared orders for food and drinks.

Worked on the behalf of 20+ different newspapers, including the Seattle Times, to both grow their customer base and retain departing customers.Topped the leaderboard numerous weeks.Sales channels included cold calling and emailing.

Gained experience working with K-12 students by tutoring mathematics, including Algebra, Geometry, and Calculus I & II.Developed problem solving skills in students to help them succeed on their own.