A collection of my recent work and side projects.

Faculty of Science Data Warehouse is a data engineering project designed to build a centralized data integration system based on Medallion Architecture. This system processes data from multiple sources through an automated ETL pipeline, stores the data in a PostgreSQL Data Mart, and presents analytical insights through an interactive Apache Superset dashboard to support data-driven decision-making

Rice phenology classification system based on Sentinel-1 imagery using a Geo-AI approach, Python, Random Forest, and Flask API, achieving 87% accuracy to support automated rice growth-stage analysis.

Automated Attendance System Using Face Recognition is a web-based attendance solution powered by Computer Vision, Dlib Facial Landmark Detection, and Dlib's ResNet-based Face Recognition Model. It performs real-time face recognition and automatically records attendance with high accuracy, enhanced security, and improved operational efficiency.

Budgetin AI is a smart financial management application that helps users record income and expenses, monitor balances in real time, manage data through Excel import/export, and ask financial questions using an AI assistant. This app is designed to make financial management easier, faster, and smarter

Integrated web-based livestock monitoring system using IoT sensors and ESP32-CAM to track ammonia, light intensity, temperature, and humidity in real time

Sistem Retrieval-Augmented Generation (RAG) untuk dokumen PDF menggunakan Google Gemini Embedding, ChromaDB, dan Groq LLM. Dilengkapi dengan evaluasi metrik dan visualisasi performa.

An image classification project that uses computer vision to recognize and classify hand gestures into rock, paper, and scissors with fast and accurate prediction results

This infographic presents Indonesia’s transition from fossil fuels to renewable energy as a strategy to achieve zero emission. It highlights the impact of fossil fuel use, renewable energy potential, emission trends, and the role of government policies in supporting a cleaner and more sustainable future

Platform ITERA Intellectual Property Management Center (PKKI) is a web-based system designed to manage, monitor, and analyze intellectual property data within ITERA. This platform supports structured management of copyrights, patents, simple patents, and other intellectual property records in an integrated and accessible way.

FINARA adalah sistem manajemen ritel dan keuangan yang dirancang untuk membantu bisnis mengelola penjualan, inventaris, supplier, hutang-piutang, dan akuntansi dalam satu platform terintegrasi. Sistem ini dibangun menggunakan Next.js dengan lebih dari 10 modul dan 19 tabel terintegrasi, serta dilengkapi fitur Balanced Entry Accounting Engine yang mampu menghasilkan jurnal otomatis dan laporan keuangan secara real-time

The model used is a CNN with regularization and data augmentation to improve generalization and reduce the risk of overfitting. The training data consists of three music genres: Ambient, Pop, and Rock, with a total of 910 audio segments. Each track is divided into 10 segments, and 13 MFCC features are extracted to represent the audio characteristics as input for the classification model. This model achieved 90% accuracy in classifying music genres.