Portfolio

Fullstack School Website Developer
Mental Health Detection & Classification
Driver Drowsiness Detection
GolekFood
RECCOFFEE
Capital Bikeshare Data Analytics
Clickbait Headline Classification
Pyspark Water Quality Classification
Disaster Tweet Detection

Fullstack School Website Developer

This school profile website showcases various school-related information and was independently developed by me, covering design, frontend, and backend. The website was built using Vue 3 and TypeScript for the frontend, Hono and Cloudflare Workers for the backend, PostgreSQL as the database, Prisma as the ORM, and Cloudinary for media management.

Live Website
Mental Health Detection & Classification

This project classifies and detects mental health conditions through social media. A framework called Symptom-aware Multimodel Mental Health Detection (SyMMHD), based on the MentalBERT Transformer model, is proposed in this project.

Private Code
Driver Drowsiness Detection

This project detects driver drowsiness in real time by integrating the system with the car’s dashboard camera. The system was built using the VGG-16 model.

Source Code
GolekFood

Runner-up of Amikom ICT Award (AMICTA) 2023 in the AI and IOT category

GolekFood is an Indonesian food or drink recommendation website based on nutritional value. As an AI Engineer and data scientist in this project, I am in charge of cleaning data, analysing data, applying machine learning algorithms and deploying models so that they can be accessed through APIs.

Live Demo

Source Code
RECCOFFEE

Best Capstone Project at Dicoding Academy Certified Independent Study X Kampus Merdeka

RECCOFFEE is a coffee bean recommendation website based on user preferences which include aroma, acid, body, flavour, and aftertaste. The recommendation system on this website is made using the Gaussian Naive Bayes algorithm.

Source Code
Capital Bikeshare Data Analytics

This project is a dataset analysis project for Capital Bikeshare, a bicycle-sharing system that serves Washington, D.C., and certain counties of the larger metropolitan area. The goal of this project is to gain insights from the data in order to answer several business questions that determine the company's decisions.

Live Demo

Source Code
Clickbait Headline Classification

This project is a clickbait news headline classification project using three different deep learning algorithms namely LSTM, Bi-LSTM, and GRU. The three algorithms are compared based on accuracy, precision, recall, and F1-score, and training time. This project is also an experiment from a research with my lecturer at the university.

Source Code
Pyspark Water Quality Classification

This is a project that aims to classify whether water is safe for consumption or not based on its chemical content. This project is run using the PySpark library.

Source Code
Disaster Tweet Detection

This project is an experimental research project with my lecturer that aims to classify whether a tweet contains disaster event information or not. This project uses the SVM algorithm by modifying some parameters in it.

Source Code