About
I started out mostly building for the web, working across interfaces, APIs, and backend systems. Over time, I became less interested in simply getting a feature to work and more interested in what sits underneath it: how data moves, how different parts of a system affect each other, and why some implementation decisions hold up better than others.
That curiosity gradually pulled me into other areas. Native development made me think differently about interaction and platform constraints, while working with machine learning made me think more about uncertainty, how to evaluate model behavior, and when a model is actually the right tool for a product.
I’m still figuring out what I want to go deeper into, and I like that part of it. My interests tend to move as I discover new things, and the projects I build usually follow that curiosity.
How I approach building
Most of my learning still happens through projects. I like starting with something I don’t fully understand, building far enough to expose the difficult parts, and then looking back at what I would keep, change, or explore further.
I don’t approach every project the same way. When the goal is to prototype, validate an idea, or ship something quickly, I’m comfortable using tools that help me move faster. But when the goal is to learn, I try not to outsource the part I actually want to understand. I’d rather slow down, work through the implementation myself, and come away knowing why it works.
I also like thinking through the direction of a project before getting too deep into implementation, especially when early decisions can shape the rest of the system. Once I’m building, I tend to pay attention to the parts that could become unreliable, unnecessarily complex, or genuinely worth optimizing.
Afterward, I like revisiting the decisions I made and asking what I would keep, what I would change, and whether the trade-offs still make sense with what I know now.
What I’m exploring now
Right now, I’m spending more time with native Apple development and ML systems, particularly retrieval, RAG, model evaluation, and workflows where LLMs interact with tools or structured data.
What interests me most is usually the part after the first demo works: whether the result is reliable, whether the right context is being used, how failure cases behave, and whether a model is actually the right solution in the first place.
For now, I’m using projects and smaller experiments to explore these ideas rather than treating any one of them as a fixed specialization.
Background
I’m currently studying Informatics Engineering at Sepuluh Nopember Institute of Technology and learning at Apple Developer Academy. Alongside that, I work part-time as a backend engineer, contributing to an ERP platform and other business systems.
I’ve also worked on client projects and public-facing systems, including a province-wide student admissions platform in East Java. These experiences have let me work with software in very different contexts, from internal business workflows to systems serving a much larger audience.
Find me on
Or mail me at hi@andeluw.my.id