Apr 2025 – Jul 2025 · 3 mos

May 2022 – Jul 2025 · 3 yrs 2 mos

• Leading multiple GPay Engineering teams in India & Singapore. Building full stack tech for lending(credit carsd, loans & more) & Monetisation.
• Eatablished and lead - all AI/ML initiatives at Google Pay since 6+yrs (including Ads, Scratch cards, Games, ML Models for credit for both consumers & merchants)

Apr 2019 – Apr 2022 · 3 yrs

• Leading ML Engineering and ML Application teams for Google Pay to make it the smartest payments app . I built, recruited the team from scratch across India and Singapore.
• I initiated, led, planned and built a charter for ML Platforms that enables personalization and ranking usecases for India, US and Singapore Gpay Users. These platforms simplify development, life-cycle management and serving of ML models for 100s of millions of users in real-time and batch modes.
• We developed several ML modesls and systems that enable smarts for key Gpay initiatives like user-acquisition and retention. User personalization on homepage by ranking and recommendations of business, people, recommending tags for transactions and identifying transaction patterns. I am currently focused on enabling financial services like loans and credit cards for all Gpay users.
• I am passionate data privacy and building novel privacy-safe ML modeling techniques and engineering systems to unlock power of data across organizations

May 2018 – Apr 2019 · 11 mos · Bangalore

• Leading storefront data science , data engineering and operational data science teams. This encompasses customer shopping experience, smart inventory management, logistics and data engineering systems

Feb 2016 – May 2018 · 2 yrs 3 mos · Bangalore

• Leading a team of 16 data scientists to deliver projects in six key business areas – Personalization, Recommendation, Returns, Fashion Design (RAPID), Merchandizing and Search
• Led research and development of personalized size recommendation system using deep embeddings, auto-encoders and GBM. Resulted in 1% reduction in returns and 2% improvement in conversion
• Ideated, researched and led development of personalized real-time re-ranking of search results based on user’s current
• in-session activity and past interactions. Resulted in 23% improvement in click depth.
• Developed and executed strategy for personalized user experience across key sections of the mobile app – feed on home page which resulted in 20% improvement in revenue per impression, ranking on list pages and recommendations on product pages which increased CTR’s by 5%.

Sep 2014 – Feb 2016 · 1 yr 5 mos

• Technical lead for research and implementation of category classification at eBay.
• Designed and developed high precision classifiers using neural networks, auto-encoders and Logistic Regression.
• Led a team to research, analyze and develop KNN based global classification system for GS1 product standards.
• Developed GBM classifier to identify spam on eBay based on image and text features.

Jan 2012 – Aug 2014 · 2 yrs 7 mos

• Customized Elastic Search module for use across multiple business units at eBay.
• Developed an algorithm for calculating similarity metric for categories based on user behavior and inventory.
• Built classifiers for toys using Logistic Regression and Clustering, resulting in better recommendations and increased sales during holiday season.

Nov 2008 – Jan 2012 · 3 yrs 2 mos

• Domain Dashboard
• Responsible for daily operation of Oversee's domain acquisition operations.
• Designed and developed domain buying methods based on collaborative filtering and improved domain retention from 2% to 18%
• Developed MVC web application to manage and control end-to-end homegrown crowd sourcing platform.
• Mined web logs, search queries and cookie information to discover features for picking keywords for online advertising campaigns.
• Data Warehouse Project
• Designed, implemented and optimized ETL process for collecting and loading stats data from existing MySql database to Netezza Data Warehouse using Syncsort tool and Perl.
• Analyzed and improved performance for hourly data loads to Data Warehouse and page loads across web server.
• Completed full cycle DW implementation, including development of ETLs, improving data models and architecture, working with several techs to rewrite current applications and setting up monitoring and backups.
• Actively led the project and interacted with developers, QA, operations, systems, product team and executives to make the project a success.
• Domain Name Appraisal Tool
• Responsible for development and maintenance of rule based appraisal engine used for evaluating domain names.
• Worked with remote team to developed XML-RPC, SOAP and web interface in JAVA to allow service to asynchronous batch processing.

Jul 2006 – Nov 2008 · 2 yrs 4 mos

• NameKing
• Owner of Oversee's registrar which handles thousands of transactions a day and is responsible for all the internal registration needs.
• Primary contact between Business, Product, Operations and Technical teams to develop new functionality and maintain day to day operations.
• Developed multi-threaded bulk processing system for registrar services capable of handling thousands of requests simultaneously
• Designed, implemented and tested feedback and rating system for users to rate transactions on live site.
• Coordinated with third party vendors and meta search engines companies to develop mutually agreed XML format for content distribution.

Dec 2004 – Jan 2006 · 1 yr 1 mo

• Preprocess ADOPT
• Designed, implemented and analyzed preprocessing dynamic programming algorithms. These techniques were exhaustively tested and results were reported in paper "Preprocessing Techniques for Accelerating the DCOP Algorithm ADOPT" published in AAMAS 2005.
• Envisioned, designed and implemented a new preprocessing technique involving storing partial solution generated by each agent and distributing them. Detailed analysis proved improvement of run time by factor of two.
• Search Bots
• Designed and developed protocols for coordination, communication and defined rules for teamwork among robots.
• Implemented Simultaneous Localization and Mapping using sensor landmarks by applying Extended Kalman Filter on the input readings

Aug 2003 – May 2004 · 9 mos

• Designed and developed in-house report generating tool that provided service agents a fast & effective way to generate various kinds of custom reports in Core JAVA and Swing.
• Heavily involved in automated test suite, documentation, quality control and testing procedures building.