Description
33219 Barmetta Ln Temecula, CA 92592 (858) 603-7852 benfactorial@iCloud.com
Ben Polovick
Highly analytical and process-oriented data scientist with in-depth knowledge of
machine learning algorithms, regression modeling, time series analysis,
probability & statistics, and big data mining tools and procedures. Excels at
furnishing business insights from big data and creating analyses, procedures and
visualizations to enable business analysts and stakeholders to make optimal
business decisions.
SKILLS
Proficient or familiar with a vast array of programming languages, concepts and technologies, including:
● Julia (Numerical Programming Language)
● Python
● SQL
● Javascript
● SAS (Statistical Analysis Software)
● Machine Learning
● Regression Modeling
● Anomaly Detection
● Big Data Mining
● Data Visualization
● Statistical Classification
EXPERIENCE
conDati, Inc. (Palo Alto, CA) — Senior Data Scientist
March 2018 – Present
● Built data visualizations, reports, and insights that helped digital marketers understand and optimize their ad spend, target audience, and digital marketing impact on the business.
● Helped conDati to build comprehensive, cohesive analytics from various data sources such as Google Analytics, Keyword Searches (Google AdWords, Bing), Social Media marketing (Facebook, Twitter, etc.), and eMail Marketing (Cordial, HubSpot, Act-On, etc.)
● Built pacing calculators and on-demand reports for a large Digital Marketer who managed social media ads for various brands. The reports provided recommendations to adjust ad spend based on desired targets such as ad clicks, website visits/page views, and budget.
● Implemented a Dynamic Time Warping (DTW) algorithm, which measures the similarity between different time series, to help with anomaly detection on marketing campaigns; useful for digital marketers who have hundreds or thousands of active campaigns and need to drill down to one (or few) that are causing spikes in ad spend or any other metric of interest.
● Implemented an algorithm utilizing the LASSO regression technique for determining which words and phrases from marketing emails’ subject line text are most impactful to open rate (which words/phrases improve or decrease the likelihood of an email open). Akamai Technologies (Santa Clara, CA) — Senior Data Scientist
April 2017 – February 2018
● Finding the needle in a haystack – Created Algorithms and Visualizations to isolate webpage assets (images, javascript, css, etc.) that most impact page load time.
● Used regression models to predict changes to websites’ page load times based on page asset response times.
SOASTA, Inc. (Mountain View, CA) — Senior Data Scientist
January 2015 – April 2017
● Helped build a library of Julia functions for big data analysis and visualization of web performance and real user activities.
● 3 patents pending:
○ Consumer Performance Index: benchmarking website performance based on the performance of top websites ○ Conversion Impact Scoring: algorithm for ranking a website’s most impactful pages, in terms of the relationship between load time and conversion.
○ Test Plan Creation: set of analyses and algorithms for building a performance/load test plan to simulate realistic user traffic.
●
●
Mentored interns and junior data scientists.
Worked directly with eCommerce and Media clients to answer tough business questions, using their RUM (real user monitoring) data.
SOASTA, Inc. (Mountain View, CA) — Performance Engineer
September 2011 – January 2015
● Consulted on performance testing design and capacity planning for eCommerce and other websites that handle huge volumes of traffic.
● Built automated test scripts for Web UI functional testing.
● Executed large-scale production load tests using SOASTA CloudTest.
● Analyzed, interpreted and generated test result reports for our customers.
● Provided performance tuning recommendations.
Intuit (San Diego, CA) — Performance Engineer
December 2007 – September 2011
● Ensured stability and reliability of Intuit’s $2 billion eCommerce platform and enterprise integration middleware.
● Administered tools for monitoring production traffic, system availability, and alerting of performance issues. Primary tools used were Wily Introscope and Wiley CEM.
U.S. Census Bureau (Suitland, MD) — Mathematical Statistician
January 2007 – December 2007
● Set up rules and procedures for data editing and anomaly detection on data received from economic surveys.
● Wrote automated SAS programs to summarize and report on nightly data received from Economic Surveys. EDUCATION
The University of Akron (Akron, OH) — M.S. Statistics
August 2005 – December 2006
Coursework Included:
● Mathematical Statistics & Probability Theory
● Experimental Design
● Design of Sample Surveys
● Response Surface Methodology
● Multivariate Statistics
● Statistical Quality Control
● Statistical Data Management
● Linear & Nonlinear Regression
● Data Imputation Methods
The University of Akron (Akron, OH) — B.S. Mathematics
August 1999 – August 2005
Coursework Included:
● Real Analysis
● Complex Analysis
● Abstract Algebra
● Combinatorics
● Linear Algebra
● Number Theory
● Differential Equations
● Intro to Cryptography
