TECHNOLOGIES

LINK

Cubic

A comprehensive payroll management application built from scratch with employee tracking, client management, and detailed analytics. Includes the web app, backend, and a mobile app. Complexities include authorization, ABAC, multi-tenancy, complex calculations, data validation, and more.

TECHNOLOGIES

LINK

AI 231 Final project

A computer vision project using deep learning and CNNs for brain tumor classification. The project involves fine-tuning a pre-trained CNN by removing the head and retraining the final layer for classification on brain tumor images. This is a full-stack project with a frontend, backend, and ML microservice.

TECHNOLOGIES

LINK

Wordle Clone

Full-stack Clone of the popular game Wordle built using Next.js. Unlike most clones which simply work by fetching a random word from a text file without taking into account the rarity of a word, this project actively tries to repeatedly fetch a random word, comparing the frequency of each word against an external API to check that it matches the difficulty set in the settings, and stops when a word matching the difficulty settings appears.

TECHNOLOGIES

LINK

Text Twist Clone

Full-stack Clone of the popular word game Text Twist using Vue on the frontend and Flask on the backend. The Flask backend works in a very similar way to the backend of my Wordle app, except written in python.

TECHNOLOGIES

LINK

Sudoku Solver Visualizer

A simple visualizer of a Sudoku solver algorithm using backtracking. The App picks a random sample problem from a set of 6 problems, and the user can visualize the solver on the press of a button.

TECHNOLOGIES

LINK

AI 221 Final Project

A computer vision project that detects species of a plant based on leaf pictures. Built using deep learning techniques to classify plant species from leaf images.