Hi, I'm Zafir
I'm a software engineer based in New York City. I received my B.S. in Computer Science from NYU in 2019 and have since focused on full stack web developer roles where I've been able to use JavaScript, Node.js, PHP, and Vue, among other technologies. After graduating, I worked as a full stack web developer for IndoorBuddy, a very early stage startup focused on bringing indoor wayfinding to grocery stores and other complex indoor venues. After IndoorBuddy, I joined Thankview as a software engineer working on both our front-end in AngularJS and Vue, and our back-end in Laravel/PHP. I have over 2 years of industry experience and am currently looking for my next role! In my free time, I've enjoyed working on projects that have allowed me to learn about topics in front-end and back-end development, databases, and machine learning. Some of those projects are listed here. You can find out more about my work from my Github and LinkedIn below.
JavaScript
CSS
Vue.js
MongoDB
HTML5
Node.js
C++
SASS
Projects
Filtos is an app that allows for advanced filtering of songs from users' Spotify libraries. Users can filter songs by genre, tempo, energy, danceability, and more and then select songs to save to a playlist. Made using Vue.js, Node.js, Express, HTML, CSS/SCSS, and Spotify's web API.
Stock Portfolio is an app that allows users to register an account with an email, log in, and buy mock stocks given $5000 cash. The app uses IEX Cloud API for current stock prices and to measure portfolio performance. Built using Vue.js, HTML, CSS, Node.js, Express, and MongoDB.
A 2 layer Fully-connected perceptron for recognizing handwritten digits from 0 to 4. The network is trained using back-propagation with stochastic gradient descent. The goal was not only to train the network to recognize handwritten digits, but to experiment with various hyper parameters of the network to determine which combination of parameters produced the most accurate network. The results of that experimentation are represented in the tables below. This neural network was created using C++.
Shooter.io is a real-time, online multiplayer game that allows users to control a circle and obtain points by defeating other users. Each player controls a circle using the left, right, up, down arrow keys or by using the W, A, S, D keys. A player can use thier mouse to point at another player on the screen and left click in order to fire a laser at the opposing player. Once a player dies, they are removed from the server and given the option to restart and play again. The game continues for the players who are still alive. Shooter.io was made using Node.js, Express, p5.js, and Socket.io.
Statify is a site that allows Spotify users to see their listening habits over time, share their data through a link, and compare with friends. Listeners can view their top artists, songs, and genres of all time, of the last 6 months, and of the last 4 weeks. A link sharing feature allows users to share their profile with friends and compare their friends habits with theirs. This app was made using Vue.js on the front-end with the Veutify component library and Node.js and Express on the backend along with MongoDB for data storage. It is currently live and can be found at the link below.
PriCoSha is a rudimentary social media site where users can create an account, comment, like, tag, post, add/delete friends, and direct message people, among other features. The main goal of this project was to display a deep understanding of SQL and database design. The site communincates with a database in MySQL in order to store and display information to the user. Tables were created for content, comments, friendgroups, members, people, sharing, and content tags. Flask for Python was used to query the database with the aim to reduce queries as much as possible to ensure the site remained fast. PriCoSha was made in HTML, CSS, Python and MySQL.
Q is a 2D platform game where the objective is to make it across the map. The players controls their block using the left and right arrow keys and the spacebar to jump. Enemies will chase after the player and shoot lasers trying to stop them. Sounds like a simple platformer, but the twist in Q is that time stops when the player stops. Time only moves forward when the player is moving. Use that to your advantage as you make your way through 3 levels of increasing difficulty. The main goal of Q was to use OpenGL to create a platformer that consisted of at least 3 levels, enemy A.I., sound effects and music, animations, and a title screen and game mode. Q was made in C++.
This project outlines a complete network infrastructure for a hypothetical architecture company called Global Design Alternatives (GDA). It includes an IP addressing schema for all offices, LAN designs for every office, the identification of every applicable protocol used in each layer of the network, the locations for the main data centers, calculations of throughout, acceptable delays, and bandwidth, and an outline for security measures.
The main goal of this project is to display a deep understanding of computer networks from the physical layer to the application layer. All information about GDA was relevant in making crucial decisions about the network infrastructure. Routing voIP traffic over an MPLS network instead of VPN through public internet, using VLANs to connect departments in an office, calculating bandwidth requirements for each department, and choosing MPLS in order to engineer traffic on the network were some of the decisions made when designing the network. The full document outline can be found here