Shweta Kumari

Shweta Kumari

Master of Science student in Computer Science at the University of Southern California

Projects

usc logo Ticketmaster Booking Application

Technologies: AngularJS, NodeJS, Typescript, GCP
  • Created a TicketMaster booking application using AngularJS for the frontend and NodeJS for the backend.
  • The app allows users to search for events using different filters, with the backend retrieving filtered results from the TicketMaster API.
  • It includes features like autocomplete and sorting for results, and users can add events to their favorites.
  • Additionally, a NodeJS API is implemented to fetch artists' data in parallel, ensuring low latency within the application.


usc logo API Documentation Tool

Technologies: React, Python, Swagger, SQLite, GitLab
  • Improved the process of creating and managing API documentation by creating an internal centralized tool using Swagger Editor, Swagger UI, and Open API Specifications.
  • This initiative effectively addressed redundancy, incomplete documentation, and difficulty locating specific documents.


usc logo Enhancement of Image Upload Architecture and Infrastructure

Technologies: Go, RabbitMQ, Redis, Docker, Containers, AWS, DynamoDB, Postgres
  • Revamped the architecture and infrastructure of the image upload service, previously taking around 400ms to complete.
  • Utilizing Golang for parallel processing, RabbitMQ for background metadata handling, DynamoDB for temporary metadata storage, and AWS for cloud-based image storage, user experience saw substantial enhancements.
  • This optimized strategy enhanced user experience during image uploads, ensuring a seamless journey within milliseconds, thereby reducing image service time from 400ms to 40ms.


usc logo Procedure for Synchronizing User Information in Go

Technologies: Go, RabbitMQ, Docker, Containers, MySQL, Postgres, Cassandra
  • The project aimed to identify and rectify unsynchronized user data across fifteen databases daily.
  • This involved implementing a scheduled cron job to identify user IDs updated within the last 24 hours.
  • Using a Golang procedure, I executed fifteen parallel goroutines to synchronize data into the respective databases based on these IDs.
  • This solution effectively mitigated the problem of data unsynchronization.


usc logo Automated GST Tax Collection, Matching, and Allocation

Technologies: PHP, Go, RabbitMQ, MYSQL, Postgres
  • To validate the GST information of potential paid sellers on Indiamart, we developed a method to enhance user authenticity.
  • It involved extracting GSTs from both a third-party API and MasterIndia API, utilizing parallel processes to identify the best-matched GST for users, and allocating verified GSTs to users.
  • This initiative resulted in a notable 20% surge in paid sellers.