I am a graduate engineer, with a Master's Degree in Computer Science. I love tinkering with new technologies, and reading about technological advancements happening around the world. Specifically those related to Human Computer Interaction, and Machine Learning in the Software Industry. I am well versed in a few Object Oriented and Scripting languages. I enjoy studying the Web and pondering over the possibilities that are yet to be uncovered in the same.
My non-technical pleasures include reading (mainly Sci-Fi, short stories and poetry), writing (the same categories), sketching (portraits, anatomical concepts), and designing (UI mocks for the web and phones). You can connect with me on my social profiles (links at the bottom).
My personal motto is Per Aspera, Ad Astra, and I try my best to live by these words. Although, I do complain sometimes.
You can know more about me, or connect with me over the following,
Building amazing things as part of the Blackfoot Command and Control team in EC2 Nucleus.
Managing, (learning) and improving how the Typography and UX behind the most widely used operating system in the world are developed
Helped on-board the new Fonts package system to the Windows 10 Store
Assisted in cleaning up older projects from the pipeline and streamlining the testing and debugging workflow
May 11, 2017
Graduated with a CGPA of 3.48
Graduated with a GPA of 4.0
Built a Java-based iteration of the 2D Arcade Shooter, Crimsonland with Processing.
Uses movement AI behaviors like Pursue, Seek, Arrive. It also uses Decision Making AI constructs like Finite State Machines and Decision Trees. Primarily an exercise in studying and evaluating popular Path-Finding algorithms like A* Search and Dijkstra.
Built an AI system that comprises of a clique of Wolves. Each tries to be a part of the pack and follows the one they deem worthy. They all have individual perception radii and their own speeds for traveling the landscape.
A Java application made with the help of Processing, a framework for geometry based animation. WolfClock was mainly made to adorn the digital visualization wall at James B. Hunt Jr. Library on campus at NCSU.
Graduated with a GPA of 4.0
Built a 3D implementation of the Arcade classic to understand how Computer Graphics are rendered. This was an extensive exercise in JavaScript-based application development, and understanding how to optimize GPU workload.
Frogger was made using Three.js a popular framework for developing graphics on the web for applications and games alike.
Assisted Dr. Robert St. Amant, in conducting research on Design Principles for handheld and mobile devices.
Study included understanding Finite State Machines in a deeper sense and finding optimizations. Implementation work involved writing formal chapters on the findings and designing an integrated web-based book.
Interned with the Innovation Council for Cyber-Security.
Configured and deployed an OpenStack system to manage all Development Projects for the division. Integrated the same with a private portal designed in Django. To serve as a way of managing access and updating permissions on finances and committee approval.
Added functionality by setting up a locally connected GitLab and OpenProject services for thorough project management. Conducted UX activities to come up with the best portal design for employees of the division.
Graduated with a GPA of 3.55
Successfully completed a series of mini-projects that deal with a variety of algorithms used for Data-Guided Business Intelligence.
Relatively larger projects involved were designing a Music Recommender System using Apache Spark, and a Twitter Sentiment Analyzer and Visualizer using Kafka and MatPlot.
Moreover, relatively complex and difficult to understand projects involved performing Time-Series analysis, Bayesian Regression on BitCoin currency values, and Word-Vector based learning and calculations.
Upgraded the User Experience and functionalities of a website designed for Middle-Schoolers and children of the age-group.
Conducted stakeholder meetings, and carried out extensive user-testing on both students and teachers. Compiled results and implemented improvements that led to smiling children and an insane selfie to summarize the experience.
Worked as a Copy Editor for The Technician, official Newspaper of NC State University.
Graduated with a GPA of 2.89
(Per Aspera, Ad Astra)
Designed a simple Python application that uses the Big5 Personality Test to determine the users' personalities.
It uses the total data to estimate the Social Comfort level of the people present and accordingly suggests discussion topics or activities everybody can participate in, without feeling outside their comfort zone.
Designed a web application targeted at Writers.
Carried out several iterations of user-testing and relied on feedbacks of several writers to come up with the best design. Used the GOMS model, Fitts' Law and Keystroke Analysis Model for designing an efficient interface.
Designed a streaming Python application that uses the Twitter Streaming API to listen to certain calls and respond to them intelligently.
It uses past data to predict current phone ringer volume based on surroundings and the reactions received by people upon having different ringer volume levels. Past data is set of tweets received and the data about reactions it contains. Simple perceptron was used to train the ringer-bot.
Let's learn to stand on our feet.
Graduated with a First Class.
Designed a box-office success prediction system for movies produced in Bollywood.
The components involved are the Front-End (a web-based portal displaying current predictions), a Listener (listens to the Twitter stream and collects most relevant tweets), a Classifier (for capturing the sentiment of the tweet collected), a Prediction-Engine (which is built on ID3 Algorithm for prediction based on feature values confined to a set).
Also wrote a technical paper describing the system.
Published a paper on IJCA for building a theoretical feed-optimization model, which tries to rank importance of posts based on the quality of the content, influenced by certain domain-experts' opinions and the implicit rating collected estimated from the average time spent reading the article.