Shivam Vaish

Experience

• 

  Samsung Electronics

  Jul 2017 - Sept 2018

  Engineer

  1) Enhanced the Smart Deadlock Module used for automatic deadlock detection and log reporting by including new functionalities for detecting deadlocks caused by mutex locks and join cases, which accounted for 40 percent of the deadlock scenarios.2) Optimized a large library for use in the Smart Deadlock module by creating a tool to remove unused code based on a report of unused functions, resulting in a 90% reduction in binary size.

• 

  Paytm

  Sept 2018 - Nov 2019

  Software Engineer

  1) Crafted a multithreading solution for the Cashier page loading flow, reducing page load time from 1 second to 300 milliseconds.2) Rolled out UPI Intent as a payment method, achieving a 30 % increase in transaction completion rates, a 25% reduction in average payment time, and a 15 % boost in user satisfaction by streamlining the payment process with pre-filled details and direct integration.3) Spearheaded the complete redesign of the SMS gateway architecture, leading to a 40% reduction in downtime and a 20% reduction in cost. Show less

• 

  Adobe

  Nov 2019 - Sept 2020

  Senior Member Of Technical Staff

  1) Built and rolled out versioning features for Adobe Lightroom, enabling collaborative photo editing and efficient version management.2) Resolved production alerts and enhanced system performance, contributing to technical discussions and design decisions that advanced backend systems and refined user workflows.

• 

  Amazon

  Sept 2020 - now

  Software development engineer 2

  1) Led the development of a high-performance, event-based system, ingesting around 8 lakh rows in 10-15 seconds with configurable inputs, efficient retries, and robust failure reporting. Oversaw Step Functions, microservices, and a team of 2.2) Migrated the Report Management System to Vibe, reducing replication time from 4 hours to 30 seconds, and led a team of 6 SDEs and 2 interns in metrics planning, authorization design, and a multi-phase migration with rollback to meet a 5-month deadline.3) Created an automated ingestion system for planning tools, accelerated cross-file validation by 40%, and established zero-downtime VPC segregation for microservices.Facilitated design reviews, sprint grooming, planning, and retrospectives for a team of 11.4) Architected a machine learning-based locker recommendation system, enabling constant order access to the top 20 recommendations, and achieving a 2% increase in purchases and a 12% boost in selecting the top access point. Show less