Hello, I'm
Pranav Rayudu

I'm a full-stack developer who loves building robots and creative programming.
Say Hi!

UT Automatic Course Notifier

UT Automatic Course Notifier

Course schedule bot to monitor, notify, and register for courses automatically.

Created general framework for a course monitor bot that can monitor courses and notify user of any changes. Backend bot written using Selenium, Flask, and Postgresql. Notifications can be configured via multiple channels, including Slack.

Frontend dashboard written using React, Redux, and Ant design. Password protected for security.

Courses and running jobs can be persistently stored in DB and resumed after restart. Configured for fully automated deployment onto Heroku.

Visit Github

Squarecap - Software Intern

  • Developed live data feature for all question types along with live student lists for production app in React and Angular.
  • Wrote and updated Selenium scripts to scrap and filter professors from over 11 universities as potential clients for Squarecap.

  • Developed live data feature for all question types along with live student lists for production app in React and Angular.
  • Wrote and updated Selenium scripts to scrap and filter professors from over 11 universities as potential clients for Squarecap.

AmigOS Operating System

Unix-style operating system with an integrated shell, inter-process communication, and text editor.

Team lead of Shell project for Operating Systems final, consisting of 11 students over 3 subteams. I worked with all teams to provide clear guidelines for the final product and oversaw code integration.

Implemented UNIX-style signals within own subteam.

Picture This

Picture This

Cross-platform, fullstack app that crowd-sources affordable attractions around town. Uses a React Native and Flask + MongoDB stack

Cross-platform, fullstack app that crowd-sources affordable attractions around town. Uses a React Native and Flask + MongoDB stack

Visit Github
Kinetic Keys

Kinetic Keys

Pose Detection game implemented in React.js and Tensorflow.js; Showcased at Convergent Side-Project Expo

A browser-based game where users act-out letters with their body to spell out words. It detects letters from key body joints through a custom pose estimation model running on Tensorflow.js. This custom vision model was refined further from the PoseNet model using a 4500+ image dataset we created for this project.
Made in collaboration with Mehul Daruka and Mayank Daruka.

Visit Site
Paragon

Paragon

Bot to monitor Slack channels for aggression using Natural Language Processing; Uses React.js and Node + MongoDB stack.

Bot to monitor Slack channels for aggression using Natural Language Processing; Uses React.js and Node + MongoDB stack.

View Github

Neural Net for MNIST dataset

Self-contained dense multilayer perceptron written in C++; trained using stochastic gradient descent

Self-contained dense multilayer perceptron written in C++; trained using stochastic gradient descent

View Github

Realtime Identity Tracking Research

Developed and benchmarked software on top YOLOv3 and Deepsort to recognize and track people; written in Python

Developed and benchmarked software on top YOLOv3 and Deepsort to recognize and track people; written in Python

Read Paper
Hurricane Hero

Hurricane Hero

Hurricane damage detector and reported implemented using React.js and Flask + CNN classifier running on Azure Cloud

Hurricanes are incredibly dangerous and powerful storms, leaving many families dislocated, and their homes destroyed. Hurricane Heroes is a program that uses a binary image classification to detect concentrations of damaged houses to help authorities decide what areas they should focus their efforts. We trained a model on Azure cloud using satellite data from the National Geodetic Survey and create a visual representation to show what areas were most damaged by a hurricane. Other team members consist of Ajith Kemisetti, Sahil Jain, Daehyun Kyoung, and Martin Nguyen.

Visit Devpost