Projects

Caching Non-Blocking API Client, Optum, July-August 2020

I worked in a team of four to prove the efficiency of a caching non-blocking API client. We built blocking, non-blocking, and caching non-blocking API clients in Java. Then, we created an API Management resource using the Azure platform, and created an API with a GET request. We used this same GET request to be processed through each API client that we built, and compared the time it takes to run through each client. We were able to show that the it takes the least amount of time for a caching non-blocking API client to process the same API request, showing that asynchronous programming is more efficient than synchronous programming. We also took our project a step further by implementing caching to serve data even faster. This project was completed as a side project within my internship at Optum for the summer of 2020. Therefore, all of my work cannot be shared openly because it is owned by Optum.

Proving the Capabilities of Airship, Optum, June-August, 2020

I worked in a team of nine to prove the capabilities of Airship, a customer engagement platform. We completed proof-of-concept work by constructing demo apps in Android and iOS using Kotlin and Swift. Specifically, I worked on the Android app, integrating different capabilities of Airship such as sending push notifications, researching the Google Fit platform, and interpreting performance analytics. Similarly to the caching non-blocking API client, this project is own by Optum and I cannot share any source code or specific information regarding the project.

Escape Game, WPI, March-May, 2020

I used Java to build different variations of Escape, a board game in which the goal is to move pieces until a player escapes from the center of the board. I used test-driven development to structure how I wrote my code. Click here to checkout my final verison of the game on GitHub, including documents that show the requirements of the project, and a written report of what I did.

Interactive Qualifying Project (IQP), WPI, January-March, 2020

As a part of WPI's Development Design Lab, I worked in a team of three to develop teaching and learning materials for the Prebysterian Primary School in Akyem Dwenase, Ghana. We used our problem-solving skills to co-design physical learning tools alongside the teachers that worked at the primary school. Furthermore, we took this project beyond its physical deliverable by establishing a solid foundation of trust and mutual learning between ourselves and the community of Dwenase. Click here to read a cover letter explaining the project and how our final papers are structured, and click here to read through my final papers. Click here to learn more about WPI's Development Design Lab!

Building Machine Learning Models for Action Classification, WPI, October-December, 2019

I worked in a team of two to conduct research and build out machine learning models for a data set of different actions performed during experimental sessions. We used Matlab's classification, regression, and clustering algorithms to determine which models worked best for the given data set. Click here to read my final paper!

Comparative Sequence Analysis of EEEV, WPI, October-December, 2019

I worked in a team of five to conduct bioinformatics research on the Eastern Equine Encephalitis virus. We used T-Coffee to perform comparative sequence analyses between EEEV sequences from different years, strains, and host species. We also performed phylogenetic analyses to show evolutionary behaviors among different strains. These analyses identified several distinctions in important amino acid residue chains. The goal of this research is to understand genetic differences and potential weaknesses in EEEV. Click here to read my final paper!

Bayesian Networks, WPI, October, 2019

I worked in a team of two to implement a data structure that represents a Bayesian Network. Bayesian Networks models conditional dependence for probability computations. The structure we built represents the nodes, edges, and the conditional probability tables (CPTs) for each node. We used Java to build out our data structure. Click here to check out my repo on GitHub!

Decision Trees for Connect Four, WPI, September, 2019

I implemented a decision tree model in Java. The model takes a connect-four board state as input and produces will win if the game is continued. I extracted five features from the raw board state given in the training data set to predict the final winners. Click here to check out my repo on GitHub, including a written report explaining more about what I did.

Application Onboarding Interface, Optum, June-August, 2019

I worked in a team of two to build out a user interface for onboarding applications into an internal company system. We used Angular7 to develop the front-end, utilized an OrientDB database for the back-end, and used serverless functions with Fn Project to serve information throughout the tech stack. I took the lead in developing the user interface and learned SemanticUI and PrimeNG to improve user experience. This work belongs to Optum. Therefore, I am unable to share more information relating to source code or documentation.

Prototype of a New User Interface for Bannerweb, WPI, May, 2019

I worked in a team of three to build a prototype of a potential re-design for the user interface of Bannerweb. Bannerweb is a software that WPI uses to manage student registration, grades, and other information. My team used Powerpoint to construct a prototype of a new user interface for Bannerweb. This prototype had a much more organized and clear flow to accessing important information as compared to the current Bannerweb software. Click here to read our final paper that contains links to our prototype!

Meeting Scheduler Application, WPI, October-December, 2018

I worked in a team of 4 to build out a meeting scheduler application using Java and AWS. I was responsible for constructing the entire front-end using vanilla HTML, CSS, and Javascript. The AWS account that our team used has expired, so the functionality of the app is no longer accessible. However, click here to read our final report!