Informatics Engineering student at Universitas Pancasila, based in Jakarta.
My work sits in two halves that I am deliberately trying to join. On one side: backend services and full stack applications, where the concerns are correctness, state, and what happens when a request fails halfway through. On the other: applied deep learning, mostly scientific machine learning, where the concerns are formulation and evidence.
Open to internship and entry level software engineering roles, particularly where engineering and applied AI meet.
| Backend |        |
| Frontend |       |
| Machine Learning |      |
| Tooling |      |
Digital wallet backend covering auth, top up, payments, peer to peer transfers and reporting. Transfers settle asynchronously through a background worker, so the request path stays fast and funds are never lost between two accounts. 26 passing tests. | Browser based markdown editor whose PDF export is driven by the browser's own print engine rather than a screenshot, with configurable paper size, margins and page break rules. Fully client side, nothing is uploaded. |
My personal site, including an authenticated admin area that manages its own content. | Kolmogorov-Arnold Networks implemented from scratch. B-spline bases built from the Cox-de Boor recursion in pure PyTorch, then benchmarked against a parameter matched MLP under controlled conditions. |
Replication of Raissi et al. (2019). Solves the forward PDE without a mesh, then recovers hidden viscosity and advection coefficients from sparse observations alone. | Offline first installable PWA for daily volleyball hitting practice. Layered service worker caching, Web Audio cues, Wake Lock, and streak tracking with export and import. No framework. |
| Project | What it is | Stack |
|---|---|---|
| Hidden Fluid Mechanics | Reconstructs an unobserved pressure field from velocity data alone, supervised only by the Navier-Stokes residual | PyTorch |
| CatatStok | Desktop inventory management application, built as primary developer | Java, JavaFX, Maven |
| Bron-Kerbosch Visualizer | Step by step maximum clique visualizer with pseudocode highlighting and a live recursion stack | Python, Pygame |
- Building an AI integrated engineering project end to end: model, inference service, dashboard, container.
- Studying Japanese, working toward JLPT certification.
- Reading around scientific machine learning and its use in industrial and manufacturing settings.