Kushagra Srivastava

Hi, I'm Kush.

I'm a final year computer science student at The University of Manchester with an interest in software engineering and FinTech, and have a major interest in high performance computing, distributed systems, and low latency programming. I have experience in software development in financial and software firms, and am constantly looking to learn more through hands-on projects. My pinned projects including a trading system created using C++, capable of listing buy and sell orders for any equity of your choice. This project is being developed consistently so if you have any pull requests or suggestions please don't hesitate to contact me. I am currently focusing on development of the trading system, as well as the compiler for my own language "Micro" that can be accessed through my GitHub.

GitHub

Data Analytics Internship

I worked on a very exciting internship at GKN Automotive as a data analyst, working with IT management teams across multiple sites globally to implement analytics dashboards, to visualise IT service desk data. I identified optimal KPIs for the IT management team and created dashboards to summarise service level agreement data. I also used correlation analysis to identify inconsistent use of the platform by engineers, through analysis of biases across the SLA data, and collaborated with management to determine the training issues causing these biases.

Software Engineering Internship

During my internship with the TornadoVM team, I worked on an exciting project to create a system that tracks how well software benchmarks perform. My main tasks included designing and creating a database from scratch, writing scripts in Python to collect and organise data, building a user-friendly dashboard to show performance data, and connecting our new system to Jenkins pipelines for automated testing. I worked with Django and Jupyter Notebooks, as well as standard frontend technologies such as HTML, CSS, and JavaScript. The project aimed to provide the developers with a framework that allowed them to store benchmark data on the Django database and compare benchmark times on the same device over time.

My Projects

Web Applications
FooBar is a social media web application created with vanilla JavaScript, HTML, CSS, and PHP. It serves as a chat, bar/restaurant review, and timetabling application, where you can coordinate meet-ups with friends at any of the bars, clubs, and restaurants in Manchester. Click to check the GitHub below. View my GitHub profile for other web application projects of mine. I enjoy experimenting with Three.js, React, and new CSS frameworks.

GitHub

Capital Discord Bot I used the Discord API and the discord.py library to create the Discord chatbot "Capital", and hosted it online using Heroku's hosting service. As a second part to this project, I designed a sorting algorithm to create a ranked list of most volatile stocks, using data web-scraped from TradingView using Beautiful Soup 4. This data was then sent to the chat when requested by a specific command. Many additional commands are available for this bot. September 2023 Update: Project now hosted through Render. Long periods of inactivity may therefore cause slow responses initially.

Invite Bot

Machine Learning Models I create machine learning models from scratch and also using libraries such as sci-kit learn. I have created models that work with facial recognition data using different statistical techniques such as linear least squares regression and gradient descent. I have also created natural language processing projects such as spam filters and text classifiers. These projects aren't publicly available but can be demonstrated on request.

Contact Me!

Email
LinkedIn
UpWork