Building India's largest premium discovery and trial platform where you can enjoy a specially curated online shopping experience hassle-free. Try Something New & Be Something New.

Digitizing & Organising India's Local Retailers to power the largest retail chain in the world & deliver the Neighbourhood Shopping Experience of the future.

Leading a team of 30 odd exceptional engineers in building the tech required for the company. Have contributed to the design, architecture and implementation of the following products.Customer Relationship ModuleHandles everything from farmer registration, profile management, order management, offers etc.Catalog ServiceManages a multilingual catalog of products, crops, diseases, solutions etc to serve the CRM, the Farmer app and several other platforms. AgroExOur own in house Last mile delivery solution for rural India. This initiative has helped up provide next day deliveries even in remote villages in Gujarat, Maharashtra and Rajasthan. It manages everything from consignments, packages, delivery partners, delivery tracking, payments, reconciliation, digital wallets etc.

Audience ingestion layer (Core AdServer Team)Designed and implemented the ingestion architecture of Audience data for Ad Serving. This involved realtime ingestion from multiple data partners which basically translated to > 1Billion requests per day. Used Mongo as the backend store and ZMQ for asynchronous ingestion.Aerospike as Audience data storeImplemented the intermediate layer for using Aerospike as a high performance audience store. This had to be done when MongoDB started breaking for the kind of load we had and we were forced to move to a faster alternative.

Corores VPL Designed and developed the first of its kind visual programming language for educational robotics. It let kids draw flow charts that could directly be used for programming a robot. Has facilities like variables, loops, functional abstraction etc. Corores Logo (CLogo) The CLogo tool is a Python based compiler that facilitates robotics programming through a LOGO like language.

Latency Regression Prevention (LRP team) While a member of the Latency Regression Prevention team I was responsible for building tools for assisting detection of latency regressions. This involved building JavaScript based interfaces that were used to automate generation of graphs and other statistical results. Developed a tool to check for latency changes across deployments. Built in Java, the tool would take two deployment numbers, analyze the latency changes between them and point out the widgets/packages responsible for it.Detail Page Website Development Worked with a team responsible for managing the widgets on amazon.com detail pages.