We decided to build Archiviette, a student attendance and marks tracking system for our class. Why? Because spreadsheets are boring, nobody maintains them properly, and also because our collective resume currently peaks at “can print Hello World” and then immediately falls off a cliff.
The team is me, Pacman, and Dino. Three people whose entire programming experience is knowing how to print "Hello, world!" in Python and feeling way too accomplished afterward. We don’t know Flask. We don’t know React. We don’t know SQLite. We don’t even know what the fuck JSON is, we just know it’s pronounced the same as Jason. And somehow, this felt like a good time to start a real project.
The Current State - Empty Folders and Delusion
Right now, Archiviette exists as a GitHub repository with four empty folders (frontend/, backend/, database/, docs/), a default README, and an MIT license we added because every serious-looking repo has one and we didn’t want to look unserious.
No code. No logic. No features. Just empty directories, misplaced confidence, and the belief that this will eventually become something usable.
Division of Labor (Mostly Guessing)
We split things up in the most logical way we could think of, which is impressive considering we barely know what we’re doing.
Pacman is handling the database, which means learning SQL, designing tables, and understanding why foreign keys exist instead of just free-balling it. Dino is on frontend duty with React, which means figuring out JSX, components, state, and why missing one bracket can destroy your entire app. I’m doing the backend with Flask, which mostly involves learning routing, APIs, and why nothing works the first time you run it.
None of us are specialists. We’re learning just enough to move forward, breaking things constantly, Googling like maniacs, and slowly piecing together how all these parts are supposed to communicate without collapsing.
The Tech Stack
We’re using React, Flask, and SQLite mostly because every beginner tutorial on the internet said they were “simple” and “beginner-friendly.” Right now, beginner-friendly just means “there are enough Stack Overflow threads that someone else has already suffered through this exact problem.”
The goal is to keep it aggressively simple. No fancy dashboards. No overengineered analytics. Just login, attendance tracking, and exam marks. If those three things work reliably without nuking the database, we’re calling it a success.
The Learning Strategy
We’re using AI tools like ChatGPT, Copilot, and Claude as explanations and guardrails, not as “do everything for me” buttons. The idea is to understand what’s happening, write most of the code ourselves, and suffer through fixing our own mistakes.
Is this the fastest way to learn? Probably not. Is it honest? Yeah. We’re not trying to speedrun competence, we’re just trying not to quit halfway through.
Git & GitHub
We know the basics: clone, add, commit, push. Everything beyond that currently feels like unexplored territory. Merge conflicts sound horrifying. Branching feels unnecessary until the exact moment it becomes mandatory. Commit messages are mostly vibes.
We are fully expecting to mess this up at least once. Possibly in a very public and educational way.
The Timeline (Very Optimistic)
Week 1: Get Flask running, render literally anything in React, and create a SQLite table without breaking it.
Week 2: Make everything talk to each other. Frontend to backend. Backend to database. Minimal screaming.
Week 3+: Start building actual features - login, records, attendance, marks.
We have absolutely no idea if this timeline is realistic. Reality will let us know very soon.
Why This Matters
Archiviette isn’t going to change the world. It probably won’t even look impressive for a while. But it’s ours. It’s the jump from “I followed tutorials” to “I built something people can actually use.”
We’re okay with being bad at this for a bit. That’s how learning actually works, even if it sucks.
What Happens Next
We’ll keep updating this as we go. When things work, we’ll write about it. When things break, and they definitely will, we’ll write about that too.
And we'll include the repo link once we have something to show. For now, it's just an empty private repo.