Yoosif Sherif

A platform for one of the world's largest companies in information technology services. This platform is used to administer managed cloud hosting on AWS and Microsoft Azure. My role involved Angular front-end app migration, code reviews, feature development, and bug fixes.

A web automation tool that allows users to supply a set of users and a set of questions and then bypasses reCAPTCHA to ask the questions. I used an external API for reCAPTCHA. The tool is smart enough not to ask a single user a question more than once and not to flood the user with all the questions it has to ask simultaneously. This way, it does not appear like a bot is performing the tasks.

This tool is also a Chrome extension that calls a database to check the status of Ask.fm users when they visit their Ask.fm profile page. For a better developer experience, I used React for the UI and tools like Chrome Extension Hot Reloader.

This tool acts like a scraper that scrapes users and saves users that fit a specific criterion in a database. It was built with jsdom, InversifyJS, and Requests. It uses various third-party APIs to help analyze the data.
(https://github.com/yoosif0/scrape-analyse-and-save-users)

This is an end-to-end Arabic text-to-speech system I developed. This project turned out to be the best open-source Arabic TTS in terms of natural speech. I also developed microservices for this project.

This is a solution that allows IoT device owners to save their data securely and grant access to their data without relying on third-party providers. It utilizes blockchain, IPFS, and a hybrid encryption approach, specifically symmetric and asymmetric. I used the Ethereum smart contract platform and React, web3.js, and Redux on the front end.

Microservices that deploy an Arabic text-to-speech system based on Tacotron neural network architecture. The system includes these four microservices:

1. An Angular app deployed on Zeit.
2. A synthesizer deployed on a server that has a GPU in Google Cloud Platform (GCP).
3. An Arabic diacritizer deployed on Heroku.
4. A controller deployed in Heroku that turns on and off the GPU server, communicates with the diacritizer, and interfaces with the front-end app.

I also used Google Cloud Storage to store machine learning models and another GPU server in GCP for training.

This is a Node.js back-end service and an Angular app that allows donors to register by clicking on their location on a map with search and location functionality. The map adds, deletes, or updates a donor's location, which is reflected in real time without the need for a page refresh for other observers, including donors and patients.

A platform that is concerned with online buying, selling, and bidding services. My role involved building Angular views, adding unit and end-to-end tests, and providing functionality through a mock API.

A research exploration I performed while earning my master's degree at The University of British Columbia (not a project). This research involved utilizing NLP to infer whether a commit was a bug fix by using metadata, such as the commit message. I used a GraphQL API.

A decentralized React app that allows patients to share their private medical data securely with selected doctors without the need for cloud storage. It depends on the blockchain with an Ethereum smart contract, IPFS, and a hybrid encryption approach, both symmetric and asymmetric. In addition to React, I used Redux on the front end for state management and web3.js for interacting with Ethereum nodes.

I provided two React courses on YouTube—one in Arabic and the other in English. The courses explain the basic concepts of React and Redux. Please refer to the GitHub URL for the links to the YouTube playlists.

A research exploration I performed during my master's degree at UBC. In this research, we wanted to determine if we could detect fixed bugs in open-source projects using static analyzers.
I automated the process of cloning GitHub repositories, performing diffs along the repository commits, extracting and saving affected files for each commit, and performing linting operations on those files.

An app that allows users with different authorization access levels to keep track of their daily caloric intake and provides valuable analytics. I used the MEAN stack in addition to an API that gets dietary data.