Jonathan Novak Email & Phone Number

Cleveland, Oh, Us

New York City, Ny, Us

San Francisco, California, Us

• Oversaw the data science department which was responsible for building and maintaining our suggested weights algorithm, dynamic weight modes, churn and engagement models, recommendation engines, experimentation framework, and data lake house infrastructure.• Navigated a post-RIF environment by developing OKRs with our product and performance teams, building career growth plans for individuals, and implementing improvements based on regular feedback from the team.

San Francisco, California, Us

• Managed a team of three data scientists who developed and improved our recommendation system for programs and workouts.• Led the development of our cloud-based infrastructure for model deployment using SageMaker and implementing MLOps best practices.

Rolling Meadows, Il, Us

• Built a brand new corporate data science department by developing a recruiting process, defining a career framework, and establishing OKRs for the team. • Managed a team of eight composed of data scientists, data analysts, and machine learning engineers while also collaborating with sales, marketing, and IT. • Developed a churn model to predict high risk clients using XGBoost which led to a 70% reduction in churn for high-risk clients within several pilot markets. • Produced a UBCF model which recommends upsell opportunities to current clients and generated $50M in expected revenue. • Deployed machine learning models through Azure Databricks and MLflow while applying software engineering best practices such as version control, unit and integration testing, monitoring, and alerting.

Chicago, Illinois, Us

• Managed a team of six that ingested various geospatial data sources such as weather to improve failure models for industrial machines.• Developed geospatial algorithms to improve joining different data sources at varying resolutions which led to a +10% model improvement.• Championed agile project management and implemented and coached several teams to reduce the amount of time to deliver milestones and build better models through an iterative and progressive process.• Led the development of the data science manager recruiting process, built a career framework for managers, and outlined management best practices for the data science department.

Rochester, New York, Us

• Managed a team of seven individuals including data scientists and data engineers that delivered data science solutions for sales and marketing.• Built an account segmentation model using K-means clustering to group accounts based on similar demographics. The clustering model was optimized using the silhouette score and the results provided our sales and marketing teams with insight into customer purchasing patterns.• Developed a SKU recommendation engine that ranks and recommends products at a store level using collaborative filtering and KNN models. This model was deployed to a custom CRM application for our sales organization to use and generated over $80M in expected revenue.• Led project team in building and automating a time series model to forecast the volume trends of internal and external products which helped determine space allocation at retail accounts. This model was built using Facebook’s Prophet package in Python and had a MAPE of less than 20%.• Performed regular code reviews, managed version control through Github, and automated testing and deployment through Circle CI. This significantly reduced debugging, troubleshooting, and time to deployment for data science models.

Washington, Dc, Us

• Researched, analyzed, and calculated the liquidity of non-equity securities for the “Liquidity and Flows of U.S. Mutual Funds” whitepaper which supported the SEC’s liquidity management rules for mutual funds and ETFs. This rule supported liquidity risk management programs for mutual funds and ETFs along with several other requirements to mitigate liquidity risk.• Conducted research on predicting corporate bankruptcy using decision trees, AdaBoost, Random Forest, and Rotational Forest models. These models provided significant performance improvement on predicting corporate bankruptcy when evaluating the models based on the AUC-ROC. • Acted as a subject matter expert in big data analytics and led projects to build relational databases in SQLite through R and Python which organized datasets from numerous vendors. This included writing programs in Python to parse hundreds of Excel spreadsheets and thousands of XML files to extract relevant information. This work allowed Economists to efficiently analyze data from several data sources which helped in rule making and research.

Rochester, New York, Us

• Developed localize pricing strategies by building predictive models and calculating elasticities through advanced multivariate statistical techniques such as multiple regression analysis. This analysis on price elasticites and cross-price elasticities identified markets which were more effected by price changes when controlling for several factors such as distribution.• Produced several ARIMA time series models for all markets to evaluate seasonal and overall trends of the company’s brand portfolio along with competitors’ brands. These models were then implemented to predict future trends along with understanding competitors’ share and pricing direction. Further, this allowed the company to determine whether competitors would increase prices in certain markets and what actions to take.• All this advanced statistical analysis was conducted in R using various packages such as time series and decision tree libraries and visually displayed in either Spotfire or Excel. Further, programs were created to efficiently analyze data for all markets along with backtesting models to determine accuracy. Finally, SQL queries for IRI data and internal databases were used to mine, extract, and then efficiently display data which usually contained +100,000 rows.

Rochester, New York, Us

• Used JDA software to adjust statistical models to reflect changes in depletions based on historical trends, new products, and price changes. This provided a more accurate forecast for production and sales and limited overproduction and out-of-stock issues.• Analyzed and tracked forecasting error and bias to determine what factors created strong deviations between forecasts and actuals.• Gathered intelligence from sales and marketing while communicating forecasting issues which allowed for more transparency throughout the company.

Rochester, New York, Us

• Managed inventory, orders, and forecasting data for spirits and wine products at bailment states to ensure product availability and prevent out-of-stock issues and penalties.• Created a complex penalty model to forecast and prevent future penalties at Pennsylvania by adjusting orders to minimize monetary costs. This model identified thousands of dollars in overstock penalties and errors in PA’s penalty structure.

Rochester, New York, Us

• Analyzed and adjusted Lewandowski and Fourier forecasting models to reflect changing patterns in depletions and shipments for numerous products which provided more accurate forecasts for production and sales.• Analyzed shipment data for products to determine warehousing usage through JDA software which lowered warehousing costs for discontinued shipment sites.

Pittsburgh, Pa, Us

• TA to two professors in Intermediate Macroeconomics and Formal Political Analysis. • Organized recitations twice a week for 20-30 students to answer questions and clear-up uncertainties which improved students’ comprehension of the material and grades in the course.• Prepared materials and taught students concepts through exercises which resulted in praise from students and professors for adequate preparation and enhanced learning for students.

New York, Ny, Us

• Organized, researched, and analyzed housing market and realtor data. • Constructed informational mortgage presentations for clients and agents to improve their knowledge of mortgages.• Organized mass quantities of data through Excel, which allowed for efficient and effective tracking of market trends.

• Researched competitors and industries for mid-cap European and African companies.• Provided analysis of companies for possible mergers and acquisitions opportunities which gained recognition by management for filtering relevant company information and providing excellent firm evaluations.• Developed an industry analysis report discussing the UK’s rum industry which provided senior management with a greater understanding of the potential market and allowed them to enter this market with confidence.

Minneapolis, Mn, Us

• Trained with financial representatives to gain experience with life insurance, investing, and financial planning for the consumer market.• Evaluated customers’ personal financial situations and provided professional advice to improve their financial stability.• Obtained Life & Health Insurance license; wrote 15 insurance policies.

Master Of Science - Ms, Data Science, Computational Methods Concentration

Master Of Science - Ms, Finance

Bachelor Of Arts - Ba, Economics And Political Science