- Led the Redirector Optimization project, reducing checkout and renewal page loading times by 150-200 ms through strategic changes, including removing unnecessary redirects. Optimized external API calls, eliminating redundant COLP and NSL calls, which further cut loading times by 1-2 seconds, enhancing user experience and platform efficiency.- Enhanced the Pricing API by implementing Redis caching within a microservice, reducing response times from 1.5-2 seconds to 50-100 milliseconds, which significantly improved performance.- Implemented a wallet feature that detects bad BINs during checkout, automatically replacing them with the next healthy card, resulting in a 20% increase in cart closure rates.- Implemented critical alerts for subscription and payment statuses in Norton customer portal, coordinating with external teams for seamless integration.- Developed CREME, a reactive library for React, enabling dynamic component changes based on rules during initial render.- Developed an Offer screen and integrated cancel reasons into the unsubscribe flow, reducing opt-out rates through A/B testing; also created a UI feature to onboard new TP-Link products into our system.

- Designed & developed an Event-based notification system, which provided a scalable solution & reduced the load of the DB.- Enabled self-service manual renewal flow for monthly subscription users, addressing the 20% of users not enrolled in automatic renewal. Post-implementation, achieved a 60% conversion rate from manual renewals, significantly enhancing customer retention and revenue generation.-Led the successful upgrade of Drools, configuring the WildFly server for seamless deployment. Developed a custom migration tool that automated the migration of missing data from an Excel input file, significantly reducing manual effort from 2-3 days to mere minutes across multiple environments. Additionally, created a rule comparison tool capable of processing 1.6 to 2 million files daily, optimizing data validation and reducing network latency from 0.5 seconds to just 5-6 ms.

• Conducted and organized workshops on various technologies within the college and motivated students to start learning technologies, also conducted motivational sessions.• Managed 1 batches of Core Team Members consisting of 10+ students.• Lead the App and Web development work of the club.• Brought up a Developer Ecosystem in my college, and it was recognised as the most active tech club.• Many juniors got impacted due to this club and they have started to learn new technologies by attending our peer-to-peer workshops..• Developed leadership and management skills.

- The internal anti virus evaluation tool's acceptance rate increased by supporting various non-standard malicious file names and allowing the user to submit the entire folder.- Developed features like add comments for the post, notification for mentioned user and unread message in threat platform to support researchers to discuss day-to-day operation.

When a disaster strikes, the World Food Programme (WFP), as well as the other humanitarian actors, need to design comprehensive emergency operations. They need to know what to bring and in which quantity. How many shelters? How many tons of food? These needs assessments are conducted by humanitarian experts, based on the first information collected, their knowledge and their experience.The mission of the WFP Innovation Accelerator is to build the world’s #1 disaster logistic prediction tool for cyclones.

• Created a deep learning model CNN with transfer learning to detect plant and weed using capture images in phone and tag location of weed to apply pesticide in future.• Gather, organize and input information into the DL model. • Image Processing and Segmentation.• Created Mobile Application using flutter and ML model hosted in AWS.