Sep 2022 – Dec 2024 · 2 yrs 3 mos

Aug 2018 – Sep 2022 · 4 yrs 1 mo · San Francisco Bay Area

• Search Ranking and Relevance.
• Published work:
• Improving Deep Learning for Airbnb Search,
• https://dl.acm.org/doi/abs/10.1145/3394486.3403333
• Aug 1, 2020, KDD '20: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining
• Managing Diversity in Airbnb Search,
• https://dl.acm.org/doi/10.1145/3394486.3403345
• Aug 1, 2020 KDD '20: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining

May 2016 – Aug 2018 · 2 yrs 3 mos · San Francisco Bay Area

• Lead consumer search team at DD
• Build a search service from scratch to production with read (runtime api) and write (data ingestion pipelines) using elastic search as reverse index, redis as cache, postgres as DB.
• Worked on building and scaling major components of search backend that power homepage recommendations, carousels, autocomplete, search for DD consumer product.
• Designed and Implemented major backend search components of homepage search, carousel, autocomplete components for Doordash consumer search
• Implemented a neural network model for store to store similarity
• Blog: https://blog.doordash.com/personalized-store-feed-with-vector-embeddings-251ad7a2c09a
• Architected and implemented personalized ranking search infrastructure for recommendation store feed
• Blog: https://blog.doordash.com/powering-search-recommendations-at-doordash-8310c5cfd88c
• Designed and implemented online road distance selection using isochrones
• Blog: https://blog.doordash.com/how-we-designed-road-distances-in-doordash-search-913ef8434099

Jun 2014 – May 2016 · 1 yr 11 mos · Palo Alto

• Relevance Search Team at Groupon.
• Designed and implemented keyword autocomplete feature backend for Groupon’s search box end to end.
• Setup offline pipeline to generate and index autocomplete keywords on search history in elasticsearch.
• Implemented elasticsearch script to rank autocomplete terms online.

Feb 2013 – May 2014 · 1 yr 3 mos · Seattle

• Part of Fulfillment optimization team responsible for designing and maintaining algorithms and models for supply demand allocation problem for customer orders, transportation problem, fulfillment center load balancing problem.
• Implemented Integer Programing based model for inventory allocation problem using XPRESS Solver engine. Inventory allocation problem can be stated as: given a set of items in a customer order(demand), and a set of fulfillment centers(FC) with a supply picture(supply), finding the optimal way to determine which items are shipped from which FC. The model was applied
• after the existing heuristic based sub optimal greedy solver to improve its solutions.
• Implemented Integer Programing based solution using XPRESS to model transportation problem where invetory needs to be transported pro-actively with in fulfillment centers, to place inventory in an optimal way close to the customer.
• Technologies: Java, Spring framework, junit testing, Mockito

Jun 2012 – Aug 2012 · 2 mos · San Jose

• Designed and implemented a fast retrieval system for read only e-commerce log data supporting 100 GB data.
• Pre-processed log data into a column store using Hadoop.
• Developed a high performance real time query execution engine over the column store.