I am taking a deliberate career break to prepare for a fully funded Master's degree abroad. My goal is to investigate Indonesian language technology in an academic environment where engineering, linguistics, and social context are treated as one problem—not three separate ones.

Why is Bahasa Indonesia still underrepresented in the global NLP landscape—and what does it take to close that gap properly?

Building Aibys from scratch made the representation gap tangible: data quality, tokenization, compute, evaluation, and cultural context all matter.

Why do communities with genuine needs still reject technology built “for” them?

A working village platform, training for 20–50 residents, and later low adoption taught me that technical delivery is not the same as meaningful adoption. Is the barrier language, interface, culture, daily relevance—or the way we define digital readiness?

Indonesian NLP · Low-resource languages · Computational sociolinguistics · Human-centered AI · Digital inclusion · Responsible AI

The full pipeline works end to end, and proof-of-concept training produced coherent Indonesian text. Full-scale training remains compute-constrained—an engineering limit that now motivates research collaboration.

Built and deployed a village digital platform, then delivered a one-week on-site digital-literacy program for 20–50 residents. The platform's later low adoption changed how I think: a technically correct system can still be socially wrong.

Guided visitors through UNIVERSUM · MENSCH · INTELLIGENZ at Indonesia's National Library and supported public engagement with interactive installations about AI, the universe, and human intelligence.

I do not want to build technology merely for communities. I want to understand how language, culture, trust, and participation shape technology with them.

Universitas Negeri Jakarta · 2022–2027
German Language Teacher Education · linguistics · pedagogy · public speaking · cross-cultural communication

Pac-Man contribution run

Contribution snake

2014  Curiosity began at the edges of systems: vulnerabilities, networks,
      failure modes, and how the machinery underneath actually worked.
        │
2018  Curiosity became construction. ByteDevCode and real users replaced
      experiments without a destination.
        │
2022  German-language education introduced a second lens: meaning,
      pedagogy, culture, and how communication succeeds—or breaks.
        │
2024  Goethe-Institut and Desa Medalsari connected technology with people.
      A platform worked technically but failed to sustain adoption.
        │
2025  Aibys began: Indonesian data, a tokenizer, a Transformer stack,
      proof-of-concept training, and a question larger than the code.
        │
2026  The ecosystem became public. I paused professional work to fight
      for the scholarship and research environment needed for the next step.
        │
NEXT  Turn accumulated questions into rigorous graduate research—and
      return the knowledge to the communities that shaped those questions.

“I learned how systems break before I learned how to build them. That is not a detour—it is the foundation.”