- Built ML time-series models to predict demand, sales, and site visits for multiple clients.- Generated decision-driving insights from user traffic data, both log-level event data and aggregated.- Utilized Redshift data warehouse and Python scripting to build deep, multi-step analytics algorithms. Automated these procedures as recurring daily/weekly/monthly deliverables.- Supervised day-to-day data operations and developed innovative solutions for new and existing questions in the AdTech world.- Clients included Fiat Chrysler, NBC, eBay, T-Mobile, Nintendo, Pizza Hut, and others.

- Developed and implemented a series of deep data analytics projects to help clients (Liberty Mutual, T-Mobile, Qantas) drive online conversions by utilizing user-level cookie logs stored in Amazon Redshift SQL data warehouse and analyzing users' impressions, clicks, and activities.- Developed a series of predictive models using R to predict cost-per-conversion for Liberty Mutual based on allocation of budget spend across a number of advertising channels.- Utilized Javascript D3 library to create custom data visualizations for multi-segment overlap and user path analysis.- Spearheaded migration of data warehouse from SQL Server to Amazon RedShift with translation of processes from SQL Server stored procedures into Python-driven PostgreSQL scripts.- Designed the common framework for querying client databases in Amazon RedShift utilizing Python's Psycopg2 library. This included designing a way to work with query parameters, leveraging abstraction and packaging to develop a common core across multiple complex queries, and creating a number of solutions for reducing maintenance efforts and limiting room for error.- Developed a series of multi-step SQL scripts to answer clients' questions and drive insights from the user data.- Developed a series of PowerShell/Batch/Python scripts to automate and streamline daily ETL operations.- Client projects included optimizing campaign reach, frequency, latency, multi-segment overlap, cross-channel lift, user path-to-conversion, creative rotation, quality of retargeting pool and others.

Lucky Spot aims to solve the growing problem of parking in densely-populated urban areas.We are a peer-to-peer sharing platform that allows residents to turn their unused parking spots into income by matching them with drivers looking for parking in the same neighborhood.