I'm a data scientist & economist who works with data to help people and businesses make better decisions.
Join LinkedIn & access Scott's full profile
Currently I'm figuring out what's next by talking to all the smart people I can and advising several seed-stage startups.
I lead teams that create value from data, build the infrastructure on which to store/query the data, and assemble the world-class teams and culture that make it all happen. My experience is very much industry-agnostic as I'm mostly interested in companies that have under-utilized data assets, or have mousetraps for building great datasets. I'm a strong believer that data science teams are best set up for success where there is tight organizational alignment with data infrastructure and BI/reporting.
Feedback I've received is that one of my differentiating skills is 1) given a dataset, tease out what products can be built or what business questions can be answered, and 2) give a business problem, identify the data needed to find a solution.
I have a wide range experience across different industries and company stages.
- NYC. lnkd.in/strata-nyc-2011
- Rx, SF: bit.ly/strata-rx-2012-video-SN
Big Data Business Forum (keynote), SF (other keynotes from Aneesh Chopra & Mike Olson)
Predictive Analytics World, Boston '12. (keynote)
Chief Data Scientist Summit, Chicago '12
Predictive Analytics Innovation. Video: lnkd.in/pia-chicago-2011
eMetrics, SF '12 (keynote). Description: lnkd.in/emetrics-keynote-desc
Predictive Analytics World, SF '12
Teradata River interview: lnkd.in/teradata-river
Information Week: bit.ly/MeetTheDataSci
KDNuggets interview: bit.ly/kdnuggets-nicholson
The Economist, 3/9/2012, on gleaning economic insights from the LinkedIn data: economist.com/node/21549948
SJ Mercury News, 6/26/2011, front page / hot talent in the Bay Area: lnkd.in/mercury-news
*LinkedIn blog posts (blog.linkedin.com/author/swnicholson/) on industry migration trends, careers for veterans, and gender differences in social networking
Advising on data science and product problems. HiQ Labs uses state-of-the-art analytical methods to deliver actionable insights to HR organizations so they can better engage and retain their employees.
● Brought bleeding-edge consumer internet data science to an industry in which these types of teams are rare
● Built world-class data science capability from scratch, including recruiting, technical tools, infrastructure, culture, roadmap, and integration into company operations.
● Traversed steep healthcare learning curve to identify high-value areas for impact from data science team.
● Worked with a large claims and clinical dataset to drive innovation to help hospitals, physicians, and patients make better decisions about healthcare.
● Deployed HIPAA-compliant and HITRUST-certified Hadoop cluster and data pipelines for 1) managing data from multiple external semi-structured sources and 2) data "lake" as a data science playground
● Reported to 2 CEOs (founding CEO + turnaround CEO) with tight alignment with Chief Architect, CIO, and COO.
● Team deployed tools using Python, d3, crossfilter, R, Hadoop, Kerberos/AD integration, SQL, scikit-learn, pandas, or whatever else we needed to get a job done.
● Projects included clinical recommendation engines (decision support), classifiers to prioritize high-priority patients and payment operational work, data visualizations, data infrastructure, answering the question, "what else can we do with all of this data?"
I lead a team of data scientists that focuses on increasing engagement in 3 key areas for LinkedIn: social communication (homepage, news feed, email), content (news, groups), and our efforts targeting college students and college applicants. Our work stretches from fundamentally understanding our members' experience on the site to building relevance algorithms for content. As data scientists our goal is to own the end-to-end analysis: we bring the creativity that asks the right questions of the data, deploy the right technical tools, are down in the details ensuring the data aren't fooling us, and take our insights and deliver executive-level presentations. Our goal as a team is to better inform product development decisions and importantly to isolate the deep insights that help us truly understand what motivates and drives our members. My role as team lead is to 1) develop, challenge, and assist my reports, 2) build and execute strategic data insights vision, and 3) help build out our team by building candidate pipelines and screening and closing candidates.
Beyond the product work that I am involved in, I am leading a data insights effort to understand the economic relevance and predictive power of the LinkedIn data. Previously at LinkedIn I worked on the monetization side of the business, where I led our subscriber churn reduction efforts by building survival models that targeted members with relevant content through email marketing campaigns.
I was the go-to person for all things related to data and algorithms: optimization, CTR prediction, real-time bidding strategies, product feature design and debugging, randomized experiments, data quality, data mining, client reporting, etc. As a PM, I had detailed end-to-end knowledge of our product, which was highly complementary to my data and optimization expertise.
My role on the business team was to provide data insights, create short- and long-term data & optimization product plans, and to take action and make recommendations based on what I saw in the data.
On the more technical side, I spearheaded the design of a real-time prediction and optimization system, in addition to building the core prediction models in R. I became a SQL (primarily MySQL and Netezza) query-writing expert, which allowed me to do deep dives on the data to diagnose problems and pull out valuable nuggets for the development of the core products and business.
I also was in charge of ordering snacks and drinks for our kitchen!
• Taught 300+ undergraduate and 100+ graduate students and received several distinguished awards
• Mentored, advised and guided 80 students in constructing applied econometric models to analyze real-world problems and to summarize the results in research papers
• Designed course structure, interacted with students, presented material and wrote exams and problem sets for the following courses:
-- Mathematics ‘Boot Camp’ for incoming Ph.D. Economics students
-- Mathematics, Economics, and Statistics for incoming non-technical Public Policy graduate students
-- Applied Econometrics for Public Policy graduate students, Stanford University
-- Introductory Econometrics, Santa Clara University
• Assisted economists in the preparation of presentations to the Bank President on important policy topics such as the effects of the debut of the Euro the Japanese banking crisis
• Provided official bank forecasts for French macroeconomic indicators during transition to the euro
• Redesigned and coded a new department intranet website and still-in-use daily data processing system that provides dynamically updated charts on the bank’s public website
• Wrote programs to collect, clean and analyze data using GAUSS, FAME, and STATA in a UNIX environment
Dissertation: "The Effects of Choice Context on Decision-Making: An Application to Voter Fatigue"
Have you ever felt exhausted from making decisions or been overwhelmed by variety at the supermarket? Using an econometric model and data from a highly novel natural experiment, my dissertation quantified the effect of "choice fatigue" in the context of voters making choices on long ballots.
For example, in California, voters who saw a proposition 10 positions further down the ballot than other voters were 1.3 percentage points more likely to vote "NO", and 0.7 percentage points more likely to not cast a vote in the contest. Interestingly, the magnitude of the "NO" effect is large enough to be within the winning margin of 8 statewide propositions in my dataset.
The latest draft of a paper based on one of my dissertation chapters can be here:
• Stanford Institute of Economic Policy Research, George P. Schultz Scholar, 2006-07
• Stanford University
-- Centennial Teaching Assistant Award, 2005
-- Department of Economics Teaching Assistant of the Year, 2004
-- 5-time recipient of the Department of Economics Outstanding Teaching Assistant Award
-- Department of Economics Fellowship, 2001-02
Over 300 million professionals are already on LinkedIn. Find who you know.
Principal Data Scientist
Data Scientist at Airbnb
VP of Data at Jawbone
Senior Data Scientist, Discovery and Personalization at Netflix
Executive Data Scientist, DataScienceCentral.com
Data Mining and Predictive Analytics Expert
Data Scientist at Fitbit
Data Scientist and Product Manager
Cofounder at Payability
VP of Product at RelateIQ
Commercial Real Estate Sales at Boise Valley Commercial Real Estate, LLC
Boise, Idaho Area
Austin, Texas Area
Regional Environmental Manager at Metalsa
Louisville, Kentucky Area
Production Manager at PrimeLending, NMLS# 298536
Orange County, California Area
Project Manager at ERP Consulting
San Francisco Bay Area