JSON-Shift

JSON-Shift is a tool designed to help users extract their desired information from websites and convert it into a structured JSON format based on user-defined attributes. By simplifying web scraping and information extraction, JSON-Shift makes it easy to transform website content into customized, machine-readable data. It leverages technologies such as Puppeteer for web scraping, Cohere for embedding, and large language models like Gemini and Groq for intelligent content processing. JSON-Shift supports customizable JSON output, allowing users to specify exactly which elements they want to extract, for example manga details from a reading website, personal information from a wiki page, or news articles from an Indonesian news site. The tool uses a combination of web scraping, vector storage, and LLM integration to filter out unnecessary content, retrieve relevant data, and produce a structured JSON output that meets user specifications. With features like intelligent content filtering and vector search for efficient similarity matching, JSON-Shift offers a robust solution for extracting and organizing web data into a usable format.

Raznime

Raznime is a cutting-edge anime streaming website developed using Next.js 14. The application leverages the App Router, incorporating server components, parallel and intercepting routes, and partial prerendering. These advanced features ensure a seamless and efficient streaming experience for users. For UI enhancements, Raznime utilizes Tailwind CSS for styling, Shadcn for component design, and Framer Motion for fluid animations, ensuring a visually appealing and responsive user interface. To learn more, visit the website directly.

Kostku

The Kostku Application is designed to assist boarding house owners in efficiently managing their businesses. It offers features such as recording boarding house assets, managing occupant data, handling payment transactions, and generating monthly income and expense reports. The frontend of this application is developed using React Native, the backend is powered by the Laravel framework, and MySQL is used for the database. I created this application as part of my thesis in 2021 to obtain a Bachelor's degree in Computer Science.

AkuBisa Admin Panel

During my internship at LPK Primedian Semarang in 2020, before the COVID-19 pandemic, I developed the AKUBISA School Admin Panel to support numeracy and literacy education for students. This comprehensive admin panel enables schools to manage various data, including teachers, classes, and students, and provides access to educational materials while recapitulating student scores based on multiple factors. The application aligns with the government's "Kampus Merdeka" program and was developed using the Laravel framework.

Social Video Downloader

The Simple Social Video Downloader is an easy-to-use application available on both Android and web platforms, designed to download videos from social media websites. Currently, it supports downloading from YouTube, Twitter, and TikTok. The web application is developed using the Next.js framework, while the Android application is built with React Native.

Food Recommendation API

The Food Recommendation API is an application developed for my Capstone project during the Bangkit 2021 program. I focused on the backend, creating a system to provide users with healthy food recommendations based on our collected data. The recommendations are generated using two methods: content-based recommendation and collaborative filtering. This application leverages TensorFlow for collaborative filtering, Scikit-learn for content-based recommendations, and the Flask framework to build the API.

Simple LiveChat

The Simple Livechat App is a personal project I developed to enhance my skills. It is a basic chat website where users can interact in real-time with other registered users. The application leverages Socket.IO for real-time communication, Express.js for the backend, and React.js for the frontend. Tailwind CSS is used for styling, ensuring a modern and responsive user interface.

Pepe Manga

Pepe Manga Frontend is a sleek and user-friendly website for reading manga, designed to deliver an immersive reading experience. This project focuses solely on the frontend, utilizing an external backend repository to fetch translated manga data in Indonesian. Developed with React and Material UI, the site ensures a responsive and modern interface.