Apr 2024 – May 2024 · 1 mo

Dec 2021 – Mar 2024 · 2 yrs 3 mos · Bengaluru, Karnataka, India

• Lead the technology and overall architecture, personally building the product
• from the ground up.
• Managed the design and development of the product from the very beginning,
• working across the entire technology stack, including microservices, website and
• data warehouse.
• Built and oversaw a team of 5 engineers.
• Successfully introduced a new personal finance product every 6 weeks, with
• the fastest launch completing an MVP in less than 24 hours to meet market
• demands.
• Contributed to several open-source projects while developing a distributed system to gather stock information from different public sources.

Feb 2020 – Dec 2021 · 1 yr 10 mos · Bengaluru, Karnataka

• Euclid:
• Euclid is an in-house clickstream analytics product that processes a scale of peak 1M QPS traffic.
• Responsible for development and scaling the product durind Sale periods.
• Tech stack used: Flink, Spark, Druid, HBase & Azkaban
• UMS:
• Designed and developed Unified Metrics Store (UMS), an OLAP processing
• engine plus datastore over Hadoop where users can configure MOLAP cubes
• over different kinds of source data from Kafka(realtime) and Hive(batch) tables.
• This helps users to consume metrics for analytics at any dimensional cuts across
• any subject areas and any number of days/month rollups and consume within
• a few seconds of latency. Some of features includes
• supports both event and entity source data types
• support for exact aggregators such as sum,count,count distinct,min,max,avg
• and probabilistic aggregators such as Theta Sketch, HLL and Count-Min
• Sketch (Top K)
• high cube cardinality of 500 million+
• handles very high traffic data (10TB data per day volume)
• roll-up across any number of days/month
• schema evolution support
• multiple subject area metrics in same cube
• automated backfill
• supports for deduplication of source data
• Tech stack used: Used Spark as computation engine, flink for real time ingestion,
• HBase as key/value and columnar store, dropwizard for micro service layer,
• Azkaban for orchestration, MySQL for metadata persistence, Ansible for HBase
• cluster deployment and maintenance
• Scaled UMS for BBD 2020 and other festive sales to process PBs of data without
• breaching the SLAs and with zero downtime.

Feb 2019 – Feb 2020 · 1 yr · Bengaluru, Karnataka

• Designed and developed real-time Checkout Pipeline for reporting live sale metrics.
• Designed and developed poll-based batch ingestion pipeline to stitch and ingest
• raw click stream data from Adobe Analytics to internal Hadoop cluster. Also
• developed UI to track the ingestion status.
• Designed and developed real-time Value Added Services pipeline to track the
• sale performance of bundled products.
• Scaled real-time pipelines for BBD 2019 and other festive sales to handle 5M
• peak QPS.

Sep 2018 – Feb 2019 · 5 mos · Bengaluru, Karnataka · On-site

• Developed an internal self-servicing Web App.
• Developed the Point of Sales (PoS) data processing and Web App for Finance

Jul 2017 – Aug 2018 · 1 yr 1 mo · Bengaluru, Karnataka · On-site

• Developed Competitive Pricing & Sentimental Analysis System to improve the
• product quality, marketing and sales strategy.
• Designed and Developed BI Apps usage analytics pipeline.
• Automated Tableau Server Migration process.
• Deployed Rserve in docker container and integrated it with Tableau Server

Jun 2017 – Jul 2017 · 1 mo · Bengaluru, Karnataka

• Developed and Deployed Cross Sell Engine.
• Designed and developed Price Band Optimization Model for feed.
• Design and Deployment of Multi-Source Query Engine, which can join and
• query data from multiple sources (MySQL, Hive)

Dec 2016 – May 2017 · 5 mos · Bengaluru, Karnataka

• Churn Prediction Model - To predict the customers retention for an e-commerce
• platform.
• Initiated and developed VInsights Engine, which provides various business insights to various business leads.
• Risk free Load Optimization Model for logistic allocation.
• Designed and deployed distributed Big Data Hadoop Infrastructure
• Built live traffic report from the user interaction data

Nov 2015 – Dec 2015 · 1 mo · Chennai, Tamil Nadu