Analytics Lead - Phil
San Francisco, CA
2019 - Present
- Led all data & analytics functions for mid-sized healthcare tech startup; promoted to position reporting directly to CEO
- Worked with CEO to analyze company priorities & provide relevant data for daily executive team meetings
- Created ETL to record & analyze company profitability on an order-by-order basis; no such data previously existed
- Was main source for all data company-wide, including for sales, product management, operations & finance
- Worked with external clients to determine data reporting needs; created custom external reports for several clients
- Implemented & maintained Looker instance for company-wide self-serve analytics
- Coded functions in Python to automatically enter data directly into Postgres/Looker from Google sheets
Sr. Analytics Manager - RepairPal
San Francisco, CA
2018 - 2019
- Used Python to automate multifaceted partner claims ingestion, shop repair matching, billing & invoicing process involving several external inputs/APIs, accounting for collection & reporting for over 50% of company revenue
- Implemented Mixpanel for front‐end data collection on web & mobile; designed custom ETL process to read millions of daily events from Mixpanel API and load this data into Google BigQuery in under 3 minutes
- Redesigned ETL processes for greater speed, loading 20 production web data tables into BigQuery in under 5 minutes
- Updated and refactored statistical car reliability analysis in Python, with new reliability scores being displayed on RepairPal website for over 300 new car models, plus API feeds built for display on partner websites
- Created several summary tables for more easy‐to‐understand self‐serve analytics, allowing users around the company to query data themselves instead of requiring time from analytics team
- Upgraded and simplified code for several existing Periscope dashboards
- Created new metrics for site/repair conversion; created Periscope dashboards & automated reports to keep track of new metrics for product & executive teams
Founder - Stockcallers
Oakland, CA
2018
- Built prediction marketplace on AWS where users make predictions on stock prices & cryptocurrency exchange rates in order to score points on a leaderboard; predictions can last from 15 minutes to 24 hours
- Created webservice to score predictions using a Monte Carlo simulation written in Python which runs 10,000 random walk exchange rate simulations in < 1 second; bolder predictions which are less likely to come true are given higher potential scores
- Designed backend API calls and ETL processes in Python to store minute-by-minute cryptocurrency prices in AWS PostgreSQL database, as well as update volatility & price trend figures every 5 minutes to feed Monte Carlo simulation
- Implemented alert system via email ensuring any errors in API calls or ETL processes are documented and sent to me
- Built frontend Javascript process which calls custom Python webservice to update stock & cryptocurrency prices displayed on the site on-the-fly every 5 seconds
- Designed live price charts using Node.js and custom Python webservice for custom graphics on prediction page
Analytics Manager - Thumbtack
San Francisco, CA
2016 - 2017
- Designed data schema, ETL processes, analytics strategy & reporting for new Thumbtack Payments product, working with backend engineers and Stripe engineers to determine correct data sources, as well as with internal finance and analytics teams to determine end user needs
- Revamped data schema, ETL processes, analytics strategy & reporting for Trust & Safety team, working with agents in Salt Lake City, backend engineers and leadership team to create correct data sources, redesign analytics and create new fraud & safety reporting to address increasing customer concerns on the Thumbtack marketplace platform
- Evaluated A/B tests of several product changes in Thumbtack marketplace, aimed at improving the experience for Thumbtack professionals
- Created scoring model for evaluating the population of 240,000 Thumbtack professionals, ranking them on their importance to the platform in three areas (quality, volume, centrality) in order to more effectively reward and retain the best professionals while also increasing activity levels among less active professionals
- Designed analytics & reporting for “Top Pro” rewards program, which created benchmarks for professionals to aspire to in order to gain a special badge; also created widget to surface this data in Salesforce for customer service agents
- Led SQL training sessions to enable self-serve reporting across the entire company using Mode Analytics, helping the Analytics team focus on larger projects and more complex analyses
- Automated population of descriptions in data schema using Python and accessing the Google BigQuery API, working with data infrastructure team to ensure accuracy and timeliness; this allowed all employees to understand data locations without taking bandwidth away from the Analytics team
Sr. Data Analyst - Lookout Mobile Security
San Francisco, CA
2014 - 2016
- Designed top-to-bottom ETL processes & analytics for several carrier partner integrations, including automated email reports in Ruby for external clients, Tableau dashboards for internal Sales managers, and invoicing for Finance dept
- In response to partner complaints about custom-built phone location portal, created ETL process to interpret billions of JSON Kafka messages, find relevant request & result pairs, and fill Hadoop cluster with performance data; created Hive-linked Tableau dashboard to track location performance broken down by various factors such as phone model
- Used Python to automate tracking data flows for multiple new and existing carrier partner accounts, saving time and allowing invoicing and user tracking to be passed off to lower level analysts
- Worked closely with non-technical members of Marketing/Product/Sales departments to facilitate user testing and marketing campaigns, measuring statistical significance, optimizing effectiveness, and communicating results
- Used forensic analysis to discover deep-rooted billing errors; worked with external partners and internal Engineering department to fix errors, refund customers if needed, and quantify discrepancies for financial reporting purposes
- Created central Tableau repository for Finance & Sales containing all company revenue figures, classifiable by any user-selected segments, including geographically, by product, by sales account, and by customer acquisition channel
- Created studies combining internal and external partner data to illustrate value of Lookout app in increasing partner customer retention; helped optimize sales strategies based on this retention information
- Promoted from Analyst to Senior Analyst after 1 year
Sr. Consultant - Urban Science
Woodcliff Lake, NJ
2012 - 2013
- Maintained SQL database and processing system for all Mercedes-Benz online sales leads from various sources, which were received in ADF format and then delivered to dealer CRM software in appropriate STAR5 format
- Created several dealer performance analysis reports for Mercedes-Benz USA executives utilizing SQL Server, SSRS, Visual Studio, Excel, PowerPoint, and Acrobat which were automatically delivered on a weekly or monthly basis
- Was responsible for monthly updates to online dashboard for reporting and visualization of sales lead analytics
- Created stored procedures in SQL Server 2008 for proper receipt, processing, and delivery of leads from new sources
- Conducted all quality control of sales leads, including de-duping, manual and automatic data scrubbing procedures, monthly reviews of sale closing rates by lead source, and testing with CRM vendors to ensure accuracy
- Conducted system testing for Subaru team in free time from M-B team, working with CRM vendors to ensure they received leads properly, displayed data correctly to salespeople, and returned properly formatted disposition data
Analyst - IHS Automotive Consulting
Northville, MI
2010 - 2012
- Led global data teams (China, Japan, UK, Germany, Brazil & US) in collection of automotive component supply chain information, then organized, scrubbed, and centralized this information for entry into internal IHS databases
- Used this data along with existing IHS database to create several custom market analyses, giving OEMs & suppliers all- encompassing forecasts of industry trends, supply chains, and future technologies for various automotive components
- Created several 50-slide “deep dive” intelligence/forecasting reports for OEM on competing models, using internal IHS database to predict market positioning, long-term product plans, new technology, CAFE effects, and sales volumes
- Reviewed corporate strategy for Tier 1 automotive supplier, using internal IHS database, regulatory forecasts, and financial data from client’s board meeting to recommend which product lines should be allocated expansion capital
- Created Advanced Powertrain Reports database for major domestic OEM to keep client updated on future engine/transmission technologies that would aid in CAFE/CO2 compliance efforts while retaining vehicle performance
- Was tasked with becoming department’s expert on all CAFE/Euro/GHG/CO2 regulations in order to be company resource for any projects relating to environmental compliance; also had minor focus on FMVSS regulations
- Created report for truck component suppliers on new 2013 Heavy-Duty Truck GHG/CO2 requirements by conducting interviews with truck OEMs and thoroughly examining legislation, EPA regulations, and new EPA enforcement method
Head of Data & Analytics - CarZen.com
New York, NY
2008 - 2010
- Co-created numerical system to objectively rate 400+ new car models in over 30 categories each using nearly 100,000 data points collected online; maintained this rating database for all current models
- Co-created complex “matchmaking” model to show web users the optimal car for their needs, wants, preferences, and pricing band based on input from short survey combined with objective scores from rating system database
- Managed 10-person data team in India to collect numerical results from online automotive reviews, which were used to populate rating system database
- Worked with web design team to create Flash application implementing survey, model, data, and results on CarZen website
Pricing Model Analyst - Capital One Financial
San Diego, CA
2004 - 2006
- As an intern, created complex mathematical model to investigate effectiveness of several variables and combinations thereof in predicting risk of credit card customers
- Analyzed risk benefits of a rising credit line strategy & optimized this over various customer segments, for an expected value increase of $30 per credit card account, or $30M annually
- Used advanced statistical methods in analyzing & redesigning manual superprime credit approval policy to create a fully automated strategy, for an expected net present value increase of $17 per auto loan, or $2M annually
- Maintained, utilized, and updated intensive Excel macro for monthly monitoring of superprime loan risk, as well as in-depth weekly analysis of market conditions & interest rates in order to set APRs for many types of auto loans
- Was responsible for forecasting auto loan volumes up to 5 years out, depending on auto market conditions
Intern - Mercedes-Benz USA
Montvale, NJ
2003
- Created online flowchart from translated German repair guide for work on problematic 2003 E-Class, reducing repair times for technicians and helping dealerships uphold high corporate standards for customer satisfaction
- Created & monitored nationwide daily quality data watch for the flagship Maybach model in response to high problem rate and high warranty repair costs, leading to increased attention from Mercedes headquarters in Stuttgart