Mar 2019 – Jul 2019 · 4 mos · Bengaluru Area, India

• ● Vulnerability Scanner App: Built App on Cohesity Platform to scan backed up VM for vulnerabilities using Tenable.

Jul 2018 – Mar 2019 · 8 mos

• ● Publisher Classification: Implemented a mechanism to assign a score to publishers based on their ad engagement, content engagement & traffic quality. Implemented a wrapper which categorizes publishers in high, medium & low quality buckets using these scores. This helps ad network compare how publishers are performing and which publishers need further investigation.

Jun 2015 – Jun 2018 · 3 yrs

• ● Page Engagement Analysis: In this project we built a pipeline that processes raw page engagement data like mouse movement, scroll, hover, touch, click events and produces useful metrics for spam analysis. The challenge was to handle the large volume of these events data. We accomplished it by storing the data in MongoDB and creating shards on document id hash. We wrote an api that takes hour & process name and splits the process to second level so that the process can run on multiple machines.
• ● Spam Analysis Dashboard: Worked on building a dashboard which provides insights into spam detection. Built the client using React.js and Redux (to maintain client state). Implemented features like compare view, saved view, watchlist, client sided filters, user management etc.
• ● Session Path Information: Worked on storing and querying session paths for a page.This information helps publishers in deciding the sequence of their webpages in order to keep the user engaged on their website. They can look at all the previous pages that led a user to a particular page and where the user is going next. It also helps them in deriving frequent paths that users take.
• ● Outdated User Agents: We built a system which provides a mechanism to populate rules that mark user agents latest & outdated. We looked at association between os, os version & browser name to browser name & browser version to mark browser versions outdated. We provided api that generates these rules for other teams. This information is in useful ad bidding process.
• ● Scroll Heatmaps: Designed and implemented scroll heatmaps at page level. A publisher can filter on date range and device type to analyze the parts of pages the users spend more time on. This project helped publishers with ad placements.

May 2014 – Jul 2014 · 2 mos · Bangalore

• Software Developer