I'm a Senior Full-stack AI/ML/Data Engineer/Scientist with 10 years across railway research, NLP, recommender systems, GenAI, etc. I've shipped from MATLAB simulators in academia all the way to LLMs running in production.
Currently I'm a Data Scientist at Avrioc Technologies in Abu Dhabi — building an AI-driven interactive sport coach, a GenAI chat platform, and managing LLM deployment with vLLM, Ray Serve, and Kubernetes.
Before that I led ML engineering at Mastersystem (Indonesia's largest ICT integrator), built recommendation systems at Tokopedia, NLP-for-law at Telkom Indonesia, and started in railway reliability at PT INKA after a postgrad at the Birmingham Centre for Railway Research, where my ATO-control paper was published on IEEE Xplore.
An always-on background service that monitors arXiv daily, synthesizes ML research papers into structured digests, and exposes those digests, plus a conversational RAG Q&A interface.
React frontend for ArXivAgent — an ML research digest that monitors arXiv daily, summarises papers, and answers questions about them.
Training spaCy NER models on Bahasa Indonesia text — dataset prep, custom labels, and model fine-tuning end-to-end.
Predicting customer offer-response on simulated Starbucks rewards-app data. Feature engineering, model comparison, and deployment notes.
Multi-label text classifier with a live Flask web app that categorizes incoming messages during disasters and visualizes the underlying corpus.
Stack Overflow Developer Survey deep-dive — what skills, salary, and education profile actually correlates with breaking into data science.
Email is fastest for new projects. LinkedIn for recruiting. The form on the right goes straight to my inbox via Formspree.