AI/ML engineer in Melbourne working on speech and language technology, with a focus on low-resource languages. I build production ASR and TTS systems by day and publish NLP research on the side.

• 🏢 Lead AI Engineer at Bookbot — on-device ASR/TTS across English, Indonesian, Swahili, and Colombian Spanish, plus LLM/RAG systems and the MLOps pipeline behind them
• 🎓 MSc in Information Technology (Artificial Intelligence), The University of Melbourne
• 🔬 Research interests: low-resource NLP, multilingual document understanding, RAG, and parameter-efficient LLM fine-tuning
• 🇮🇩 Part of LazarusNLP, where we train Indonesian and regional language models
• 🤗 Models and datasets on Hugging Face
• 💬 Happy to talk about speech, low-resource MT, or anything NLP — reach me by email or on LinkedIn

• MUDIDI: A Two-Stage Framework for Multilingual Dictionary Digitization with Language Models — EMNLP 2026 (first author) paper · code · dataset · docs
• Context Volume Drives Performance: Tackling Domain Shift in Extremely Low-Resource Translation via RAG — LoResMT @ EACL 2026
• NusaBERT: Multilingual and Multicultural IndoBERT — COLING 2025 code · models
• Fine-Tuning Large Language Models for Multi-Label Emotion Classification via Sentence-Label Pairing — SemEval-2025 (1st place, Sundanese track) code