Sep 2020 – Jan 2023 · 2 yrs 4 mos · Greater London, England, United Kingdom

• Instagram Reels and Camera
• Started new Instagram Reels team in London, continuously scoped projects, planned tasks, did design reviews, set up agile practices and social events.
• Drove technical decisions and consensus with all the stakeholders while continuously prioritizing most impactful projects for reels such as the new creative tool platform on IG.
• Continuously provided mentorship to various people across the company.
• Proposed and built a project for company’s hackathon related to reels which received very positive feedback from company leadership including CFO, CTO, VPs of engineering at Meta.
• US Patent pending: https://patentcenter.uspto.gov/applications/63436366?application= with Application No. 63/436,366, entitled “GENERATING AND PUBLISHING
• TRANSACTIONAL CONTENT ITEMS BASED ON ORIGINAL
• CONTENT ITEMS”
• Instagram Shopping
• Product Display Page (PDP)
• Worked on figuring the best strategy on how to improve the carousel experience on PDP to ensure users find it meaningful, engaging and ultimately help them make a purchase decision.
• Led and executed around 18 experiments which resulted in increasing time spent massively.
• Started various design and craft related initiatives in the org to improve the overall quality of shops.
• Visual Search Camera
• Created a new highly performant camera on Instagram iOS – Visual Search. Users can use to search for similar items and shop.
• Designed the whole camera architecture end-to-end and independently lead collaboration with various ML teams. Feature was later announced by Mark Zuckerberg and was picked up by many external media.
• In-Store Shopping Experiences
• Designed and implemented QR scanning and barcode scanning features for IG shops on both iOS and Android while independently leading various collaborations with cross-continent IG teams.

Jan 2019 – Sep 2020 · 1 yr 8 mos

• ▪ Experimented with various algorithms and techniques to run ML models efficiently on a mobile device.
• ▪ Implemented end to end pipeline to deploy ML model on Acrobat iOS app.
• ▪ Worked on Dynamic RNN models on Tensorflow Lite (CPU and GPU-Experimental) to compute performance metrics.
• ▪ Worked with Google Engineers on evaluating their experimental platform.

Aug 2017 – Jan 2019 · 1 yr 5 mos

• Worked on end to end development of Acrobat iOS app. Implemented many features like Create PDF, Export PDF, Combine PDFs, Social Sign-in, File System Authentication, File Operations like Rename, Delete, Duplicate, Move a file to a different folder, Copy a file etc.

Feb 2016 – Aug 2017 · 1 yr 6 mos

• Data Platform
• Worked on writing an entire end-to-end service to discovery resources from Azure in Java.
• POCs to onboard Adobe’s Marketing Cloud solutions to the Data Platform.
• Adobe Experience Manager-Communities
• Worked on Adobe Experience Manager(AEM) communities end to end development.
• Implemented social sharing, notifications, creating and updating groups to communities.

May 2015 – Aug 2015 · 3 mos · Sunnyvale, California

• Implemented edit profile feature on the LinkedIn Job search iOS app. It was deployed for millions of users.
• Automated the UI testing of all screens of LinkedIn Job Search android app using JUnit and espresso framework.

May 2014 – Aug 2014 · 3 mos · Los Angeles, California, United States

• Developed ‘Mediation optimization software’ which drastically increased advertisements effect in revenue.
• Programmed in JAVA under Linux environment, individual work.
• Processed the historical results of the advertisements campaigns using JDBC and MySQL, which turns out to be very effective in determining the most loyal, consistent and those publishers which provides the most compensation for the sale of advertisement impressions

Jun 2013 – Aug 2013 · 2 mos · Tokyo, Japan

• Developed image filtering and quality enhancement system using Fourier fringe analysis for images obtained from Mach-Zehnder interferometry experiment under Prof. Kimiya Komurasaki.
• System comprised of three steps (coded in matlab), converting raw image to frequency domain using 2D FFT, applying various filters (e.g.: Butterworth) and doing Phase unwrapping using Goldstein unwrap2D.

Aug 2012 – Nov 2012 · 3 mos · Japan

• Designing and Optimization of Spacecraft using MOEA based on decomposition for the DESTINY mission of JAXA by developing a code in C and C++ language.
• Optimized the trajectory of a space vehicle to reach the orbit at an altitude of 20,000km, minimized both the Ion engine operation time and time of flight.
• Proved that the rate of getting the optimal Pareto fronts in case of MOEA/D is faster than their previous algorithm (NSGA II) for that specific trajectory problem. That is, MOEA/D can give optimal solutions even for the lesser number of generations.