Sep 2021 – Jan 2023 · 1 yr 4 mos · Bengaluru, Karnataka, India · On-site

• At Twitter, I spearheaded efforts within the Growth team to revolutionise how users discover and engage with content, significantly improving user retention and engagement metrics. My leadership in the first week's user journey, combined with strategic hiring and innovative mentoring, marked a period of substantial growth and enhancement in user experience.
• Core Achievements:
• Growth Activation Signals: Enhanced the new user experience by leveraging crowd-sourced signals for personalized content recommendations across various surfaces. My initiatives led to increases in 14-day average weighted engagement (+0.62%), Profile Page UAM (+224k), and mDAU (+73k).
• Real-Time Recommendations: Implemented a system using real-time browsing history to improve new user recommendations, resulting in a +0.42% rise in engaging follows within 12 days.
• GDPR-Compliant Signal Transformation: Pioneered the adaptation of out-of-platform activity signals for GDPR compliance, significantly enhancing content recommendations and predicting a +1M increase in UAM.
• Global Participation: Tailored content discovery for Indian users, leading to a 2% increase in topic follows and enhanced engagement, by promoting relatable accounts over mainstream celebrities.
• Customer Journey Onboarding: Improved new user onboarding by introducing language selection options, supporting 1.2 million daily signups and boosting completion rates.
• Leadership and Development:
• Team Leadership: Led the development of user engagement strategies for days 1-7, focusing on both implicit and explicit signals to enhance early user experiences.
• Hiring and Scaling: Actively involved in the hiring process, successfully scaling the team from 3 to 10 members, enabling broader and more impactful project scopes.
• Mentorship: Mentored a team to automate the generation of topics of interest for specific regions by crawling Wikipedia and other open sources, fostering innovation and automation in content curation.

Feb 2020 – Aug 2021 · 1 yr 6 mos · Sunnyvale, California, United States

• At Google, I advanced cloud infrastructure through key projects on vSphere and Anthos On Prem, focusing on Kubernetes orchestration and hypervisor enhancements for improved system efficiency and scalability.
• Key Contributions:
• Enhanced Visibility and Reliability: I developed a sophisticated application for vSphere that enabled the tracking and alerting of resource contention at the hardware level, a critical metric previously obscured from Kubernetes' view. This tool significantly improved our insight into customer environments, enhancing the reliability of our products and services. It also contributed to a substantial reduction in the resolution time for on-call support tickets, bolstering customer satisfaction.
• Auto-Scaling Excellence: My work transcended departmental boundaries, fostering collaboration across networking, logging, and monitoring teams within the Anthos ecosystem. Together, we conducted comprehensive benchmarks of Anthos components, identifying and implementing essential scaling mechanisms. This initiative allowed the system to support an increased load of 2.5 times the original node capacity without compromising on reliability, marking a significant milestone in our scaling capabilities.
• Innovations in Autoscaling:
• Horizontal Autoscaler Integration: I led the integration of a node horizontal autoscaler for customer-facing clusters. This feature automated the scaling process based on the application demand, ensuring optimal resource allocation and system performance.
• Vertical Autoresizing Leadership: Perhaps most notably, I spearheaded the design and implementation of a node vertical autoresizing feature for admin clusters. This pioneering solution reduced our total minimum resource footprint by an impressive 50%, concurrently lowering operational costs. This initiative not only enhanced our system's efficiency but also set new industry standards for cost-effective infrastructure management.

Jun 2019 – Aug 2019 · 2 mos · Greater New York City Area

• At JP Morgan, I spearheaded the development of cutting-edge models to revolutionize financial analytics and pricing strategies. I leveraged my expertise in optimization techniques and machine learning to significantly enhance the company's trading capabilities and operational efficiency.
• Innovated in Financial Modeling: Developed a sophisticated model to calibrate implied volatility surfaces for American options, utilizing advanced optimization techniques such as Particle Swarm Optimization, Genetic Algorithm, and statistical methods. This work set new standards in accuracy and efficiency for our trading operations.
• Advanced Pricing Strategies with AI: Pioneered a hybrid model combining clustering, boosted trees, and neural networks to predict the prices of exotic derivatives. My approach reduced the quoting time from 6 seconds to near-instantaneous, dramatically improving our market responsiveness and client satisfaction.
• Optimized Machine Learning Processes: Implemented active learning techniques to intelligently select training data, achieving a 30% reduction in prediction error. This initiative not only enhanced the accuracy of our models but also streamlined our data processing workflow.
• Technical Proficiency: My role demanded a deep understanding of machine learning, optimization, and statistical analysis. I demonstrated strong technical leadership in applying these skills to solve complex financial problems, driving forward our analytics capabilities.

Apr 2018 – Jul 2018 · 3 mos · Bengaluru Area, India

• Developed the SKU platform, autocomplete and suggestions for user app and cataloguing team using NodeJS, MongoDB, Firebase, eliminating the need of user intervention post-order confirmation.
• Devised a spatial trending search for user app, increasing the user registration to task conversion ratio.
• As a part of the data team build a part of the pipeline for data preprocessing, developed a unified location storage, auto-verification and deduplication system using Django, PostgresSQL and Redis. This system was the backbone of store search, listing & SKU platform.

Apr 2017 – Mar 2018 · 11 mos

• Gopigeon is a logistic broker company for intelligent order processing, tracking and management of the whole lifecycle of retail orders for mid level ecommerce business
• Intelligent Order processing System:
• Developed a distributed bulk or API order processing system which could handle 1000 orders/min and responsible for a revenue of upto 2cr INR/month.
• Led 3 engineers to build an adaptive logistic partner allocation system using tracking history as well as user preferences of cost, service and reliability parameters. This increased the delivery by 8.9% for Indian e-commerce clients like Voonik, Lenskart, Byjus etc
• Last mile delivery:
• Led a team of 4 developers to build a component to proactively coordinated between customers and delivery agents to address the last-mile delivery issues like unreachable number, address not found, etc. This reduced the average return rate by 3% among major Indian retailers like Lenskart, Shopclues, Xiaomi, LimeRoad etc over a period of 6 months.
• Order tracking systems:
• Implemented a distributed adaptive tracker and reconciliation system to get order status through API, scraping (with captcha breaking) and manual communications for major Indian couriers like Bluedart, FedEx, Indiapost etc. This system was responsible for tracking 1 lakh open shipments/day.

Dec 2015 – Apr 2017 · 1 yr 4 mos

• Built and maintained internal admin panel dashboard to edit, place, reassign, cancel etc. client orders which was used by customer support team.
• Developed an asynchronous distributed system to process real-time orders in Django, Rabbitmq and celery and helped scale the system using AWS EC2, RDS, SQS, SNS, ALB, LAMBDA to handle 1000 orders per min with minimal downtime.

Jul 2015 – Dec 2015 · 5 mos · Hyderabad Area, India

• Developed Desflow, an internal JIRA application for communication, keeping track of errors and reference them in future for debugging, system upgrades request logging, securing confidential discussions etc

May 2014 – Jul 2014 · 2 mos · Hyderabad

• Title: Build and Deploy Portal
• Environment: Asp.Net MVC(C#), Kendo UI Reporting Tool, Telerik Reporting, Bootstrap
• Company: Microsoft India R&D Pvt. Ltd.
• Description: Build and Deploy portal generates escalation, task, build, drop reports from CMDB and TFS Warehouse DB. Automate human work to generate reports for tracking failures in build process and statistics about build process
• Title: Battleship Game for Xbox Kinect using Ripple SDK
• Environment: C#, Web API, MEF
• Company: Microsoft India R&D Pvt. Ltd.
• Description: Backend API for classic Battleship (board game) is designed with four levels of AI algorithm, each level being closer to human thinking. The back end and front end are merged using Web API.