Feb 2024 – Present · 2 yrs 3 mos · Stockholm County, Sweden · On-site

Sep 2021 – May 2023 · 1 yr 8 mos · Hyderabad, Telangana, India

• As a senior member of the Google Cloud Storage Capacity Planning team, my primary responsibilities include owning and delivering team OKRs, improving team’s efficiency, and reducing technical debt.
• Designed and implemented resilience and fault-tolerance mechanisms for the workflow, which brought down failure rates by 90%. This ensured that a daily output was generated by the system, which allowed the downstream processing to run unhindered.
• Continuing the above, I designed and implemented changes aimed at reducing manual effort involved in processing the output, by adding a second phase to the workflow, which generated a more accurate though less frequent result as a result of additional data validations mechanisms.
• I led the ownership transfer of the codebase and other software assets related to Google Cloud Storage resource planning. This involved extensive planning that covered migrations, testing and roll-back strategies, and collaboration with engineers outside my team (and timezone).
• I organised and led my team in Google Cloud’s “Fixit” weeks, where team-members volunteer to focus and work on clearing tech-debt. Me and my team members were awarded peer bonuses for the efforts.
• I identified and logged several tech-debt and refactoring tasks, which were later cleared during Fixit weeks. This significantly improved the overall quality of the team’s codebase.

Feb 2021 – Sep 2021 · 7 mos · Hybrid

• As a Software Architect, my primary responsibilities included improving the software infrastructure in terms of efficiency and reliability, and reducing technical debt.
• I made several system-design contributions to the platform, for both creating new, and improving existing systems.
• Was the go-to person for Redis and caching related topics in the organisation. I collaborated with SRE to improve the platform’s Redis deployment on Elasticache, and made contributions that improved the caching system’s resilience and availability.
• Reduced API latency by 60% and cut Redis memory usage by over 50% through a redesign of the mission-critical configurations framework. After proactively identifying performance issues with the system, I proposed the redesign to engineering leadership, and implemented it, including a new library for simplified configuration access and modification, eliminating duplicate logic and improving code quality.
• Developed an in-house caching framework using MySQL and Redis that provides APIs through higher-order functions, abstracting away low-level implementation details. This framework greatly reduced the amount of boilerplate code required, and is now widely adopted as the primary caching solution for the platform's services.
• Designed the “Approvals framework” for the platform as an event-driven state machine that worked with entities shared between tenants on the platform. I guided software engineers in the implementation. The framework was built using MongoDB and Redis.

Nov 2017 – Jan 2021 · 3 yrs 2 mos · Hybrid

• I was the lead engineer of one of the teams that built a new generation of features for Turvo, using distributed architectural patterns and strict code quality standards.
• I proposed high level design, and led the implementation of several new services. The systems were designed to be deployed separately from the legacy stack, and followed microservices patterns. I was involved in developing the microservices components, as well as starter code that all services were built on top of.
• I mentored engineers on topics such as software architecture, clean code, etc., and integrated tools such as Google Java Style maven-plugin, Sonar, etc., to the codebase to improve code quality.
• I developed the initial version of Turvo's automated code quality inspection tool using SonarQube, Jenkins and Bitbucket. The tool was designed to be triggered automatically when pull requests are opened, and then perform static code analysis, and post results back to the VCS as inline comments.

Jul 2016 – Oct 2017 · 1 yr 3 mos · Hyderabad, Telangana, India

• After Lookup was acquired by NowFloats, I led the team for NowFloats’ enterprise conversation platform, ana.chat. I designed and implemented a completely distributed, microservices-based architecture for the platform, that could be deployed both as a standalone single-tenant setup, or a multi-tenant cloud setup on AWS. The cloud setup used CloudFormation to create resources and deploy the application stack. The services were built using Java (Spring Cloud), Python, NodeJS, and Apache Thrift.
• I created a cross-functional team of web, mobile and back-end engineers from within the organisation for the project.

Jul 2015 – Oct 2017 · 2 yrs 3 mos · Bengaluru Area, India

• I proposed, planned and executed the re-architecture of Lookup platform from the initial monolithic stack to a distributed architecture deployed on AWS. The re-architecture eliminated downtimes and scaled the platform to handle more than a hundred thousand concurrent users.
• I designed the new architecture making use of AWS services such as Application Load Balancers, RDS, Elasticache, managed Elasticsearch, SQS, and Elastic Beanstalk.
• Introduced tools like Git and Jenkins to enable continuous integration and improve development process.
• Built and organised the engineering team; introduced Jira to the organisation, and set up an agile-based engineering process.

Jan 2015 – Jul 2015 · 6 mos · Bengaluru, Karnataka, India

• I was at Twitter very briefly, after the ZipDial acquisition. I continued working on the ZipDial products that Twitter kept running, before leaving the company to join Lookup.

Sep 2013 – Jun 2015 · 1 yr 9 mos · Bengaluru Area, India

• I handled development of voice-based applications using PHP and cloud telephony, which was ZipDial's flagship service.
• I created a bespoke resource monitoring and alerting tool that periodically scraped the cloud telephony platform's web page, generating actionable data. The tool enabled us to proactively take action to prevent downtimes due to capacity issues, resulting in significantly improved SLAs.
• Created a re-usable internal framework for creating cloud telephony based apps, that cut down development times by half.
• After the acquisition of ZipDial by Twitter, I continued handling the cloud telephony vertical.

Jun 2009 – Aug 2013 · 4 yrs 2 mos · Bangalore

• Innoz was founded while I was in college, where I started my career. I wrote the some of the first lines of code for the award-winning product, "SMS Gyan", and later scaled the PHP & MySQL stack to a distributed polyglot platform that handled over a million user queries daily.
• I designed and implemented the distributed design on AWS, and directly worked on setting up services such as ELBs, RDS, Elasticache, and Elastic Beanstalk.
• Built and organised the engineering team into product development, R&D, platform and support teams, and streamlined the engineering process. Hired engineers and team leads with various skill sets for all teams. Managed and mentored freshers and lateral hires.