Auto Magazine Begins: MVP, Wayfinder Map, and First Investigations

I started the first day of Auto Magazine. The project involves automating the transformation of Word documents and images into fully laid-out magazines that are ready for printing.

The MVP Scope

Today’s session helped me define the minimum viable product (MVP). This is the starting point at which the product already performs its main function, even if it does not have all the features I would like. We agreed that the result will be a system that runs in a Docker environment (a container that makes the software run the same way on any computer). The tool will take folders containing articles, use artificial intelligence to understand the structure of the files, and generate a final print-ready file with the necessary crop marks and color quality.

I have left everything that is not essential out of this initial phase. This means there will be no on-screen visual editors, complex user systems, or automated advertising management. The priority is to make the basic process work from start to finish.

Wayfinder Map and Investigations Underway

To organize the work, I created a wayfinder map, a visual guide where I mark pending tasks and their status. I currently have seven open tickets and three investigations underway. These investigations will test whether the technology I want to use will perform well before I write the final code.

The three areas I am analyzing are:

1. The rendering engine (how to lay out the page).
2. Generating the PDF for print.
3. Extracting data from Word documents using AI.

Until I have the results, I cannot move forward with the algorithm that decides where each page goes. In the meantime, my job is to prepare the test material: a set of real-world files that will be used to test the system when it is ready.

I also spent some time creating a glossary in the project’s CONTEXT.md file. This file collects definitions of the terms we will use, such as "Hotfolder" (the folder where files are placed to be processed) or "print PDF". Naming things from day one saves me from misunderstandings later on as the project grows.

As for the technology, I have options such as Astro, React, FastAPI, or SQLite on the table. I will not make a final decision until the investigations tell me which combination works best. Working this way and validating every step helps me avoid wasting time on tools that later turn out not to be useful.

The next step is to get the results of the first three investigations so I can begin structuring the system’s engine.