An interactive full-stack dashboard for Segment Anything Model 3 (SAM3), providing a complete pipeline for image segmentation, model fine-tuning, data labeling, and performance evaluation on geospatial/satellite imagery.
Frontend web app for a family-run catering, snack, and drink business. It showcases the business, menu items, customer testimonials, and provides contact information for orders.
Deep learning model training pipeline for delineating land from satellite imagery. Features preprocessing, BsiNet model implementation, and custom loss functions.
A comprehensive Next.js application designed to visualize geospatial data related to fertilizer distribution and requirements. It serves as a powerful dashboard for monitoring and analyzing data across different administrative levels (Province, Regency, District) in Indonesia.
A production-grade, standardized backend service combining Geospatial Dashboarding and Identity and Access Management (IAM). Built with Go (Fiber v3) and Clean Architecture principles, this service is designed for high performance, scalability, and maintainability.
Predicted crop yield using satellite imagery (Sentinel-2) and ground truth data. Applied classical ML models (RF, XGBoost, LightGBM) and deep learning models (LSTM, GRU) with transfer learning.
Web app turning real-time weather into absurd, funny one-liners. Collected data via Open-Meteo API, generated text with Meta-LLama-3-8B Instruct.
A simple API inspired by “No as a Service” that serves categorized and random pantun (poems). Built with Express.js.
Implemented a Zero-Shot Classification pipeline for Indonesian tweets. Preprocessed text (cleaning, slang normalization), translated to English using Deep Translator, and classified into 8 categories (e.g., Politics, Economy) using the tasksource/deberta-small-long-nli model.
Chatbot that answers questions from uploaded PDF documents. Implemented RAG with Chroma DB and LaMini-Flan-T5 model.
Analyzed Twitter (X) and YouTube comments from the 2024 Vice Presidential debate to extract sentiment and topic trends.
CNN model to classify anemia from labeled eye-bag images. Performed EDA, image preprocessing, and model training.
Interactive web app prototype for a tourism village. Designed UI components and consistent design system in Figma.
