About

I’m Wahyu Fajri Widagdo. People close to me call me Edo.

I build backend systems — Java and Spring Boot at work, Go and Laravel in the things I ship myself. I studied informatics, graduated into the middle of a pandemic, and started my career in banking, which turned out to be an unusually good place to learn how software behaves when it genuinely cannot fail.

What I actually do

Most of my work sits behind the interface: APIs, data models, migrations, and the logic that has to stay correct when a transaction is cancelled halfway through. For two years I built the customer pre-acquisition microservices at BTPN Syariah; since 2024 I have been building the test automation for those same systems.

That second half changed how I write code more than anything else has. Once you have spent a couple of years trying to break services you used to maintain, you stop designing for the happy path — you start designing for how the thing will be proven wrong.

These days I also build and ship things of my own. MeetHR is the one that is live: a career preparation platform where jobseekers get an AI-assisted CV review and then book a real session with an HR professional. I built the product end to end, alongside a co-founder who leads the business side.

KALA POS is the larger codebase — a point of sale system that keeps working when the shop’s internet does not. It is feature-complete and tested, and I’m currently packaging it into something other developers can buy and build on.

How I work

I care more about decisions than about tools. Which database, which framework, which language — those matter less than whether you can explain why the system is shaped the way it is, and what happens when the assumption behind that shape turns out to be wrong.

I use AI every day as a working tool. But the architecture, the technical decisions, and the responsibility for the code stay mine. That distinction is the whole point — reasoning is the part you cannot delegate, and it’s the part I’d rather be judged on.

I also build with language models, not just alongside them. The CV analysis in MeetHR is a production feature: it scores a CV against ATS heuristics, matches it against a specific job posting, breaks it down section by section, and predicts the interview questions likely to follow — with a paid unlock behind it and real users on the other side.

Writing the prompt turned out to be the easy part. The actual engineering was everything around it: grounding the output in the document that was really uploaded, handling the cases where the model is confidently wrong, keeping the cost per analysis low enough that the price still makes sense, and deciding what happens when the API is down and someone has already paid.

Outside the code

I write. Some of it is technical; a lot of it is not. There are older essays here about moving cities, about minimalism, about money — written when I was in my early twenties and figuring things out loudly. I’ve left them up, because pretending you arrived fully formed is its own kind of dishonesty.

A note on language: this site is in English, but much of the writing is in Indonesian — that’s simply the language I think in when I’m being honest. I don’t plan to translate the older pieces.

Say hello

I’m reachable at wahyuwidagdo31@gmail.com , and findable on GitHub and LinkedIn .

If you read something here that was useful — or wrong — I’d genuinely like to know.