Apr 2024 – Mar 2025 · 11 mos

• Working on columnar data store format.

Apr 2022 – Apr 2024 · 2 yrs

• Joined Google Core as a key member of the sessionization data processing infra team that efficiently generates sessions exceeding O(Petabytes) in size for various Google products, including Ads and Search. Focused on enhancing and maintaining the infrastructure in large scale distributed systems.
• Worked on an assignment aimed at resolving observability cost challenges for a sister team and the whole org. Collaborated with multiple clients to understand the issue and then designed and implemented a generic telemetry framework to demonstrate my ability to analyze and address complex issues effectively.

Jan 2020 – Mar 2022 · 2 yrs 2 mos

• Contributor to Google Ad's search team, focused on tackling challenges like modeling conversions that are lost due to various privacy rules. My role involved collaborating with data scientists for development of statistical models and engaging in analysis.
• Eng lead from the search side for unifying the infrastructure for sampling and reporting across multiple modeling products . Helped develop a common interface and led the search model integration end to end. It required a lot of collaboration, leadership and data analysis.

Aug 2018 – Dec 2020 · 2 yrs 4 mos

• Foundational member of the Identity Sync team at Google Cloud. Contributed to the design and implementation of sync products, building distributed systems on top of Google's DBMS and messaging queue. Gained proficiency in open-source protocols like SCIM and LDAP.
• Over the course of 2.5 years, I earned a promotion and played a mentoring role in growing the team from 2 to 12 members.

Jul 2017 – Aug 2018 · 1 yr 1 mo · Bengaluru Area, India

• Part of the team responsible to build Video Ads' division. My work expands from writing the code that is required to serve Video Ads', more particularly enhancing its scalability, modularity and reusability to building tools to enhance and extend functionalities of its backend.
• Responsible for building Access Management tool for managing access of business partners to different categories of ad creatives.
• Refactored codebase into a loosely coupled architecture​ separating business logic , video Player and adPlugin in such a way so that video Player and adPlugin can be used as extensions.

Jun 2016 – Jul 2016 · 1 mo · Bengaluru Area, India

• Built a tool to categorise delivery feedbacks into positive, negative and neutral by sentiment score generated per feedback through NBSVM algorithm.
• Built a website using python(flask framework) which displays the feedbacks at hub- level, city-level, zonal-level and provide ranks for hubs and agents, compare hubs, agents, state, city etc based on the percentage and type of feedbacks that they receive.
• Language Used :- Python, HTML.
• Libraries Used :- Jquery, Liblinear, LibSVM and NLTK.
• Frame WorksUsed :- Flask and TwitterBootstrap.
• RDBMS Used :- MySQL.