Jul 2023 – Mar 2024 · 8 mos · Bengaluru, Karnataka, India · Hybrid

• 1. Spearheaded the cloud cost optimisation initiative which includes migration of services from AWS to Azure and reducing resource utilisations. Saved cloud cost of 500K USD per annum after the migration of core services.
• 2. Technologies used: java, couchbase, redis, azure, micronaut, kafka, azure, aws & kubernetes.

Apr 2022 – Jun 2023 · 1 yr 2 mos · Bengaluru, Karnataka, India · On-site

• Lead the development of Catalog Service, Procurement Service and Replenishment Tool. All the below mentioned projects are mainstream systems and being lead parallelly.
• Key projects -
• 1. Replenishment Tool : Built it from scratch. This tool generates the POs & TOs in realtime as well as on-demand by executing various complex business rules and predictions from multiple Machine Learning Models for all the categories such as FnV, Perishables and Non Perishables. Owned the HLD, LLD and development. Lead a team of 3 devs.
• Impact - The wastages have been reduced by 10% and the entire process of raising PO/TO at store level has been made completely automated and increased the operational efficiency with no errors.
• Tech stack used - google cloud, bigquery, golang, airflow, dataflow and postgres.
• 2. Procurement System - Lead a team of 6 devs.
• Responsibility - To review PRDs with product mangers, TRD(HLD + LLD) with devs. To make the system highly available, resilient and writing code for critical modules such google cloud spanner CDC pipeline using dataflow. Making sure there are no data-leaks and change feed events get transmitted to data pipeline in realtime. Developing load testing modules for critical workflows.
• Tech stack used - google cloud, bigquery, golang, airflow, dataflow, postgres and google cloud spanner DB.
• 3. Catalog System - Lead a team of 7 devs.
• Responsibility - To review PRDs with product mangers, TRD(HLD + LLD) with devs. To make the system highly available, resilient and writing code for critical modules such google cloud spanner CDC pipeline using dataflow. Making sure there are no data-leaks and change feed events get transmitted to data pipeline in realtime. Developing load testing modules for critical workflows.
• Tech stack used - google cloud, bigquery, golang, airflow, dataflow, postgres and google cloud spanner DB.

Mar 2019 – Apr 2022 · 3 yrs 1 mo · Bangalore · On-site

• Working on distributed databases to scale out the data platform.
• I have had the opportunity to work on multiple key projects -
• 1. Communications Platform(CX)
• 2. Walmart Pharmacy(WP)
• 3. WNCP Programme
• 4. Data movement from Oracle to ADLS
• Tech stack used - Microsoft Azure, Kafka, ADLS, Cosmos DB, Redis and Elasticsearch.

Feb 2018 – Feb 2019 · 1 yr · Bengaluru Area, India

• 1. I have developed Cab Proximity Alert Engine from scratch for shared category cabs.
• (Performance stats: Throughput ~2k events per second)
• Technologies used: Apache Flink, Kafka and HDFS.
• 2. I have worked on reducing the error rates on Ola share service and making the service scalable.
• (Performance stats: Throughput ~50k/min)
• Technologies used Java Dropwizard, Elasticsearch, Kafka, Mysql and Redis.
• 3. I had worked on improving the efficiency of Foodpanda(acquired by OLA) delivery systems.

Aug 2017 – Feb 2018 · 6 mos · Bengaluru Area, India

• I worked in the core flights team and I worked on the following :
• 1. Built multi-city search functionality.
• 2. Built nearby airports search functionality.

Aug 2014 – Aug 2017 · 3 yrs · Bengaluru Area, India

• I've worked on various projects in Paytm marketplace.
• 1. Fulfillment Engine : An intelligent system, the heart of marketplace which supports seller.paytm.com and other services such as Recharges, Bus Tickets/Flights and Movies. I worked on scaling various order processing and state transition modules. Worked on adding retry mechanisms backed by RabbitMQ for the failed requests. Also, worked on making sync processing, the async way.
• 2. Warehouse Management Engine : The core of logistics. Developed from scratch, which
• gives all the possibilities of warehouses with selection scores based on the customer pin code, product attributes and seller attributes. Performance stats : 30k-50k requests per minute with CPU latency 40-60 ms.
• 3. Ticketing service : The part of travel marketplace. Working to handle huge traffic which ranges between 100k and 150k requests/minute. Also, introducing retry mechanisms for failed requests to make async processing.
• 4. Built a PoviderServiceEngine in Java to off-load CPU intensive executions (such as : text-processing, XML parsing) from nodejs to Java. Used Jetty, Jersey and Jackson.

Aug 2012 – Nov 2012 · 3 mos · Jaipur Area, India

• Evaluating and Assisting students for lab experimentation in Algorithms.

May 2012 – Jul 2012 · 2 mos · Jodhpur

• Developed Device Driver for android tablet Multi-touch(5 touches at a time) screen.
• Key Logger for linux kernel.