May 2021 – Jun 2022 · 1 yr 1 mo · Salt Lake City Metropolitan Area

• Now acquired by Pinterest, THE YES is a next generation AI-powered shopping platform poised to rewrite the architecture of e-commerce. Founded by Julie Bornstein and Amit Aggarwal, THE YES raised over $30 million in venture funding from Kirsten Green at Forerunner Ventures, Jon Callaghan at True Ventures, and Tony Florence at NEA. Other investors include Comcast Ventures and Bain Capital Ventures.
• Led growth engineering for THE YES. GMV grew >20x in one year.
• Architected and built notification pipelines and services launching new content on weekly basis. Revenue attributed to email/notifications grew >10x.
• Launched promotions and loyalty programs, email/phone capture, new methods of payments and new feed for ads among other projects; resulting in >6x jump in conversion rate overall.

Oct 2014 – May 2021 · 6 yrs 7 mos · Mountain View, New York

• Led Next Billion Users (NBU) program for Geo/Local vertical on Google Assistant. Launched 11 projects in less than a year increasing traffic for Local vertical on Google Assistant by an unprecedented 1% overall.
• Invented a new paradigm for developing human computer dialogs which is now the default framework for all dialog development on the Google Assistant. Published patent: US10803860B2: Dependency graph conversation modeling for use in conducting human-to- computer dialog sessions with a computer-implemented automated assistant.
• Designed, led development and launched infra and features on Actions on Google platform which supported critical third party APIs (e.g Helpers API, Transaction API) as well as several internal product flows. Led integration of these APIs with DialogFlow and led third party dialog quality efforts.
• Built the Knowledge graph feature which enabled task completion on Google Search (e.g booking appointments) - scaled upto 50 partners. It was Google Search’s first successful foray into the "Task Engine" which ultimately transformed into Google Assistant and was adopted into other Google products.
• Created an android app based ARCore experience quality tool which improved testing coverage by 300% and greatly reduced time to certify third party devices for ARCore compatibility. Prototyped running AR on cloud by streaming sensor data and images and running AR algorithms on the server.

Jun 2013 – Oct 2014 · 1 yr 4 mos · San Francisco Bay Area

• Startup developing the next-generation search engine for mobile devices powered by a knowledge graph of millions of real-world entities with a focus on rapidly changing data. It raised a successful Series-A of $8M+ from Sequoia Capital & Khosla Ventures.
• Architected and developed a real-time crawling, indexing and reconciliation engine.Wrote the initial batch crawler and reconciliation engine and iterated with a real-time engine which enabled fresh search results to appear at crawl speed.
• Incorporated device data such as contacts and media using the real-time engine. Integrated signals such as popularity, recency, ratings and location of entities into the ranking engine.
• Reduced the server-side query latency by over 60% to 300ms by rearchitecting search stack. Led the experiments to deploy and optimize Hadoop and HBase on SSDs, with different (and much smaller) network configurations, instead of using a large cluster. Managed maintenance and performance of backend infrastructure such as hadoop and hbase.

Oct 2010 – May 2013 · 2 yrs 7 mos

Aug 2008 – Oct 2010 · 2 yrs 2 mos

• Analytic Applications, Business Intelligence, Server Technologies, Fusion Middleware

May 2007 – Jul 2007 · 2 mos

• Woked on database performance and scalability for project assistment
• https://www.assistments.org/

May 2006 – Jul 2006 · 2 mos