6/3/24 - 8/9/24
I worked in-person as a full-time, paid computer science intern for 40 hours per week for 10 weeks in Plano, TX. This was my third internship at Capital One. I enjoyed my experience.
With four other interns, I worked on and deployed the Contribution Readiness Project, an augmentation of the CoDev website. CoDev is a platform for innersourcing - opensourcing within a company - at Capital One. The project focused on two problems. In order to contribute to an innersource repository, contributors would need to manually search through a repo to find its contribution model - the instructions and details about contributing. Examples of items included within a contribution model is a link to developer setup documentation and a promise that pull requests (PR) would be commented on within 3 days. Also, a repo's team may not keep the repo's contribution model and their contribution practice statistics in alignment. For example, a repo's contribution model may promise a PR comment time of 3 days, while in reality, the PR comment time is 5 days. The Contribution Readiness solved these two problems by displaying an innersourced repo's contribution model, the contribution statistics of the repo's team, and suggestions on how to improve the contribution model.
I used Node.js as our runtime environment to run our code, TypeScript as our programming language to reduce bugs, Express.js as our web application framework, Zod to make sure variables had the correct format, and Jest to create unit tests. After extensive review of my code by senior developers, it was deployed to the CoDev website.
Aside from development, I enjoyed participating in a hackathon with an interdisciplinary team (technology, product management, and cyber security interns) to design Business Finder. Business Finder allows large companies to find and patronize small black and brown businesses within Virginia, helping the small companies by bringing them new customers. I also enjoyed attending Capital One events (ex. TopGolf, Dominion Amusement Park, Bowling) and spending time with my friendly manager and intern group.
6/6/23 - 8/12/23
I worked in-person as a full-time, paid computer science intern for 40 hours per week for 10 weeks in Plano, TX. This was my second internship at Capital One. I fully enjoyed my internship.
I worked with four other interns on InfluxGen, which aimed to solve the following problem. The Marketing Navigator team wanted a tool that allowed them to develop data pipelines for any distributed processing framework (Apache Spark, Apache Flink, Google Cloud Dataflow Runner, etc.) without code (using only a JSON file). Unfortunately, the tool that they had only allowed them to develop data pipelines for Apache Spark without code. Consequently, the team could not make use of the different advantages that each distributed processing framework provided. The goal of InfluxGen was to develop a tool that allowed the team to develop data pipelines without code for any distributed processing framework. During my internship, I made use of Apache Beam which enabled the development of data pipelines for any distributed processing framework with code and Java as the programming language.
Outside of work, I enjoyed participating in a company hackathon. Here, I worked with another intern to develop a Node.js app that helped developers record code snippets alongside output while debugging, allowing programmers to debug without forgetting previous findings. I also enjoyed exercising and playing basketball with others at the Capital One gym and swimming at the Frisco Athletic center.
6/6/22 - 8/12/22
I worked in-person as a full-time, paid computer science intern for 40 hours per week for 10 weeks in Plano, TX. I really enjoyed my experience.
I worked with three other interns on the Solution Catalog project. This project was inspired by the following problem. Business analysts (BA) and product owners (PO) would not be able to independently view and edit the eligibility rules of different capital one solutions. Instead, they would need to ask software engineers to perform these tasks for them using SQL scripts. The goal of this project was to develop a website that enabled BAs and POs to view and edit these rules themselves.
Throughout my internship, I worked on the frontend side of the Solution Catalog project. I used Node.js as our runtime environment to run our code, Lit Element for developing web components to reduce code, used Web Pack as our bundler to reduce load time, used Type Script as our programming language to reduce bugs, and used Axios as our http client in order to allow our frontend to communicate with out backend. I also used Web Test Runner which made use of Playwright and Sinon.js for developing unit tests.
Apart from development, I had lots of fun attending different Capital One events and meeting other Capital One interns and employees. I especially enjoyed our final intern celebration in which all the interns were flown out to the headquarters of Capital One in Mclean, Virginia.
6/1/21 - 8/20/21
I worked remotely as a full-time, paid computer science intern for 40 hours per week for 12 weeks. Overall, I had a really great experience. I met a alot of different types of people, including a consultant, product manager, and technical lead. I also worked on the following projects:
myPKFiT Isitup project
I developed a Node JS based tool that can check the login page status of the myPKFiT site for over 50 countries every 15 minutes. Before this tool, the myPKFiT team would only know about login issues after their users had already experienced them. My program allowed the team to know about and fix issues before users experience them, providing a better site experience for users.
HCP Specification Documentation
Along with a fellow intern, I developed a document containing the functionalities and formatting data of the myPKFiT web application for over 50 countries. To reduce human error in data collection, I developed a Node JS based tool that automatically extracted the formatting data for each country's web application and placed the data into an excel file. This document was presented to the myPKFiT quality team and now resides in the myPKFiT technical file.
myPKFiT Screenshot Automation Tool
I developed a tool that can automatically extract screenshots from the myPKFiT site and add them to a word document. This tool was developed to remove the need for manually collecting and adding screenshots to user manuals, a process that took approximately 3 hours per country. While my tool was only a proof of concept, once fully extended by future developers, the screenshot collection process will take 10-20 minutes and be fully automated.