Jul 2021 – Jan 2022 · 6 mos

• Lead the efforts in Designing & Developing PLA(Products Listing Ads) functionally for Ads Platform in FCC. Took development & execution ownership of Ads-Serving along with that.
• Designed & developed ProductCatalogue for FCC to power multiple products like FCC Ads Manager, Pricing Manager & Fraud Detection.
• https://www.flipkartcommercecloud.com/

Mar 2020 – Jul 2021 · 1 yr 4 mos

• FCC (Flipkart Commerce Cloud - https://www.flipkartcommercecloud.com/)
• Designed & Developed Ads Attribution Reporting system for Ads performance Analysis for FCC, scalable upto TBs of data size with complex joins on Millions of customers view/click stream data.
• Enabled Data Ingestion in FCC systems by Clients by Designing & Developing Batch ingestion system for FCC Commons using Azure Storage Accounts.
• Enabled timely launch of FCC Ads Manager by Playing key role in Design, Development, e2e testing.

Jul 2018 – Mar 2020 · 1 yr 8 mos

• TNS (Trust & Safety -> Fraud Detection using Big Data and Machine Learning)
• Reduced Rating & Review Fraud by implementing micro-services for R&R fraud Models & Data pipelines for Graph algorithm processing.
• Scaled Hadoop/HBase/Yarn/Kafka by load testing & optimising configurations for Big Billion Days Sale.
• Enabled Hot & Zero loss migration & backup of HBase Systems from 1.0 to 2.0 by setting up HBase 2.0 cluster & migrating data & services with a systematic planning & execution.

Jan 2016 – Jun 2018 · 2 yrs 5 mos

• TNS (Trust & Safety -> Fraud Detection using Big Data and Machine Learning)
• Standardised TNS's Feature Generation, Feature Discovery & Feature Freshness monitoring by developing Feature Catalog and Feature Loader Systems.
• Mitigated reseller fraud & reduced cost by Multi million Dollars by implementing ML model execution micro-service and developing Data Ingestion & processing pipelines System with HBase & Spark.
• User-Insights (User Data Intelligence with HDFS, MR, Spark, HBase, Kafka)
• Unified Batch & Real-Time processing components in Athena (unified processing platform). Implemented Batch processing components joining 20+ HBase tables within few minutes using Scala & Apache Spark.
• Added intelligence by analysing & implementing multiple application pipelines like Segmentation, Affinity, Predictive models (Gender, Income, Age).

Jul 2014 – Dec 2015 · 1 yr 5 mos

• User-Insights (User Data Intelligence with HDFS, MR, Spark, HBase, Kafka)
• Scaled HBase & removed data skewness by implementing key hashing, pre-splitting, time partitioned & auto creation of HBase tables.
• Made flipkart user's location aware by designing & implementing Location Intelligence based on reverse-geocoding, IP Location with browse & purchase data with Mongo DB spatial indexes.
• POC with Titan Graph DB to generate Connection Graph of Flipkart's users.

Jul 2012 – Jun 2014 · 1 yr 11 mos · New Delhi Area, India

• worked as a research assistant on a Mobile Healthcare project in collaboration with Dartmouth university.