I'm Anant Vijay, a software developer that likes to be referred to by my moniker "DRAGON"
I am a Full Stack Web Developer that loves tinkering with Artificial Intelligence and Machine Learning and am also a DevOps enthusiast
I'm also an avid Open Source contributor, write Technical Blogs and give talks at various events about my journey and mentor newcomers
I'm well versed in the technical domains of AI/ML, Web Development and DevOps.
I'm willing to collaborate together on Open Source projects, am also interested in mentoring people that need help in tech and also like to give technical talks at various tech events, communities and conferences.
Feel free to reach out to me in case of any internship or job opportunites as well. Would also be willing to work as a freelancer too.
Application created using the T3 tech stack which uses the tRPC protocol to develope type safe APIs using TypeScript. The application allows multiple users to log into and vote amongst themselves to determine what the best quote is.
Created a Deep Learning model that can convert grayscale images to their colored counterparts. Utilized Deep Convolutional Conditional Generative Adversarial Networks (DC-GAN) in order to achieve this. Applied various GAN optimizations to improve model efficiency, training time and required data by over 80%.
Application to predict if the provided news is true or false Utilizes BERT transformer model with accuracy of 98% to identify fake news headings Consists of an user friendly UI which is also efficient at the same time.
The proposed solution for the GDSC Solution Challenge 2022, which aims to end world hunger by allowing easy sharing of food that can be consumed by others. Utilizes a CNN model with accuracy of 90% to identify food items. Consists of an attractive UI which is aimed at improving user access.
A basic game recommendation system based on non modeling approach. Used basic concept of NLP in order to implement the entire system from scratch. Highly efficient and accurate.
Augmented Reality sudoku solver that provides realtime solving of sudoku. Trained a Convolutional Neural Network with 99 percent accuracy to interpret digits. Used OpenCV for image processing. Implemented sudoku solver using backtracking.
Gathers all the internships from numerous sites at a single location for ease of access. Used scheduled web scraping and automation. Handled and maintained database for making API calls.
Application to predict if the provided news is true or false UtilizesGym exercise log for users to track their activities. Created front end using React, styled using Bootstrap, MongoDB for database and Express for backend and API, hosted on netlify.
Provides suggestions to user based on the image provided by the user. Trained a K Nearest Neighbour Model with 86 percent accuracy as well a Convolutional Neural Network with 88 percent accuracy in order to achieve this.
Maze solver, Maze/Level creator, Drawing and colouring board, Colour matcher game, Snake game, Top down shooter game, Sudoku solver, Guessing game, Hangman game, Pong game, Rock paper scissor game Created using algorithms such as A Star Path Finding Algorithm and Backtracking.