End to end pipeline chatbot built using knowledge from PDF and CSV (Bahasa Indonesia)

• Natural Language Processing

Pawductivity is a mobile app to boost user productivity. We combine a productive app with some gamification feature so user get a best experience while using this app. This mobile application built with flutter with bloc state management and retrofit while back-end is bult using Go with PostgreSQL as the database.

• Flutter
• Jenkins
• Go
• PostgreSQL
• Bloc
• Retrofit
• GORM

A static website designed to provide users with detailed information about the Pawductivity mobile app. Built using Next.js for fast performance and styled with Tailwind CSS for a modern, responsive design.

• Next JS
• TailwindCSS
• Jenkins

A mobile application to track user's expense and income, user only need to create a transaction and description and ML model will automatically classify the transaction into 5 category (Food, Home, Health, Transportation, Other). Build using tensorflow, Kotlin (Jetpack Compose), BERT (For text embedding).

• Kotlin
• GCP
• Tensorflow
• BERT

A web application designed to streamline and digitalize clinical trial management, minimizing human error throughout the trial process. The front-end is developed using Next.js, while the back-end is powered by .NET 8. PostgreSQL serves as the database, with Entity Framework utilized as the Object-Relational Mapping (ORM) tool to facilitate efficient data management.

• C#
• .NET
• Next JS
• Gitlab CI
• PostgreSQL
• Entity Framework

A comprehensive notebook demonstrating the development of a regression model for predicting housing prices. The project includes Exploratory Data Analysis (EDA), feature engineering, and model training using TensorFlow, showcasing practical applications of Machine Learning Regression techniques.

• Python
• ML Regression
• Tensorflow
• EDA

A detailed analysis of mobile phone datasets, focusing on uncovering patterns, trends, and insights through Exploratory Data Analysis (EDA). This project leverages Python to preprocess, visualize, and interpret data, providing valuable insights for decision-making and further machine learning applications.

• Python
• EDA

Built a Feedforward Neural Network (FFNN) from scratch using Python, implementing forward/backward propagation and applying it to a classification task. No deep learning frameworks were used.

• Python
• Machine Learning

Created an Artificial Neural Network (ANN) from scratch in Python, featuring customizable layers and activation functions. Used for solving a regression problem, demonstrating core neural network principles.

• Python
• Machine Learning

A web application for villa reservations based on location, price, and amenities, developed using NextJS, ExpressJS, MaterialTailwind, and MongoDB.

• NextJS
• ExpressJS
• MaterialTailwind
• MongoDB

Implemented K-Nearest Neighbors (KNN) and Naive Bayes algorithms from scratch using Python. Applied these models to classification tasks, demonstrating their effectiveness in machine learning without relying on pre-built libraries.

• Python
• Machine Learning

A web application to help college students find scholarships and admission information, consisting of a Single Page Application for users, a Monolithic interface for admins, and SOAP for integration.

• React JS
• Chakra UI
• PostgreSQL
• Node JS
• Vanilla PHP
• MariaDB
• Java SOAP

A web application for monitoring CCTV cameras, developed using React JS, Bootstrap, and Node JS.

• React JS
• Bootstrap
• Node JS
• Express JS
• PostgreSQL

A mobile application to predict skin lesion using image recognition artificial intelligence. Built using Tensorflow and CNN to classify image and flutter for the mobile application.

• Flutter
• CNN
• Classification