Designed and developed the Universal search microservice from scratch for Postman and improved product adoption by 44%.
Researched and implemented an optimized spell correction in search by customizing the Symspell library and Symmetric Delete spelling correction algorithm having p99 of <50ms.
Built end-to-end search personalization features like recent searches & recent views resulting in 5x faster discovery & navigation of Postman entities.
Developed Query Understanding Service to improve search relevance by 52% using user's data, advanced ML models, and NER using Spacy, Python, Numpy, Scikit-learn & Pandas.
Enhanced test coverage by 94% and search service reliability.
Lead and mentored a team of 4 engineers and 2 interns.