Building reliable and efficient intelligent systems for scientific and real-world applications.
I am a Computer Science graduate from the Department of Informatics, Universitas Diponegoro, with interests in machine learning research, data science, and software engineering.
My work focuses on transforming research ideas into reproducible systems, particularly in:
- Spatio-temporal graph learning
- Time-series forecasting
- Reliable and uncertainty-aware machine learning
- Computer vision for public safety
- Cross-lingual and low-resource NLP
- Efficient AI for resource-constrained environments
I also have experience in competitive data science, mathematical problem-solving, smart-contract development, and mobile application development.
A spatio-temporal deep-learning framework combining Graph Attention Networks and LSTM models to predict microseismic magnitude sequences.
The research incorporates deep ensembles and post-hoc calibration to produce more reliable uncertainty intervals across multiple temporal resolutions.
Co-authored research introducing a lightweight hierarchical time-series forecasting architecture based on depthwise separable convolution.
The model was developed for accurate multi-horizon residential-load forecasting while requiring fewer parameters than computationally intensive Transformer and LSTM baselines.
Published in Advance Sustainable Science, Engineering and Technology, 2026.
-
Gold Medal — ICECInno 2024 AquaGuard: Real-Time Drowning Detection System Using Computer Vision for Enhanced Child Safety University Undergraduate Category, Universiti Malaysia Pahang Al-Sultan Abdullah.
-
2nd Place — HOLOGY 7.0 Data Mining Competition, 2024 National technology competition organized by the Faculty of Computer Science, Universitas Brawijaya.
-
3rd Place — Data Slayer 3.0 Machine Learning Competition, 2025 National data-science competition organized by Telkom University Purwokerto and joined by 163 teams.
-
Universitas Diponegoro Representative — ON-MIPA Mathematics, 2023 Selected to represent UNDIP at the Central Java regional stage.
A research-oriented framework for multi-step microseismic magnitude prediction using spatial graph attention, temporal sequence modeling, and calibrated uncertainty estimation.
PyTorch · PyTorch Geometric · GAT · LSTM · Deep Ensembles
A real-time driver-drowsiness detection system supporting browser-camera and video-based inference.
The application detects conditions such as focused driving, talking, yawning, and microsleep, then generates an audio warning when potentially dangerous behavior is detected.
PyTorch · MobileNetV3 · LSTM · ONNX · Streamlit · WebRTC
A cross-lingual information-retrieval system for searching Javanese Wikipedia articles using Indonesian queries.
The project explores both lexical and neural retrieval through TF-IDF, multilingual BERT, LaBSE, and fine-tuned language models.
NLP · Information Retrieval · mBERT · LaBSE · Streamlit
A smart-contract system for transparent and accountable government-budget disbursement.
It models budget allocation, payment proposals, vendor validation, multisignature approval, and fund disbursement through Soroban smart contracts.
Rust · Soroban · Stellar · Smart Contracts
A modular LaTeX thesis template for Informatics students at Universitas Diponegoro.
It includes APA 7th-edition references, automated numbering, and an organized multi-file document structure.
LaTeX · BibLaTeX · Academic Tooling
- Reliable graph neural networks for scientific forecasting
- Calibrated uncertainty estimation
- Lightweight architectures for time-series applications
- AI systems for public safety and social impact
- Reproducible machine-learning experimentation
Statistics are generated from public GitHub repositories and may not include private work or contributions made to repositories owned by organizations and other users.
I am open to research collaboration and software projects involving:
- Scientific and applied machine learning
- Graph neural networks and time-series forecasting
- Computer vision and intelligent safety systems
- Indonesian and regional-language NLP
- Efficient and uncertainty-aware AI
- Open-source research software
For collaboration or professional opportunities, reach me through LinkedIn or email.