Information Systems graduate from ITS Surabaya. I build data-driven systems across fraud detection, NLP, and recommendation — with a particular interest in turning machine learning experiments into practical applications.

Most recently, I completed an undergraduate thesis on fraud detection using an autoencoder + LightGBM + Bayesian optimization pipeline.

Project	Description
ecommerce_fraud_detection	Undergraduate thesis project on the IEEE-CIS fraud dataset — combining an autoencoder for anomaly scoring, LightGBM for classification, and Bayesian optimization for pipeline tuning.
nlp_project_sriwijayaair	Sentiment analysis on airline customer reviews.
scent2me-rec-service	Content-based perfume recommender using TF-IDF, served through FastAPI.
slev	Local-first personal finance tracker built with Expo + React Native.

Fraud detection · anomaly detection · ML pipelines

My recent work has focused on taking machine learning from experimentation toward practical, end-to-end systems — from data preprocessing and feature engineering to modeling, optimization, and serving.

_{building somewhere between data, systems, and curiosity.}