So, after a long time of denial and strong regret for taking so long, I finally opened myself to the AI fever — and I loved it! In my mother language we have a saying for this: "I bit my tongue."
Working on my famous Sushi Project, I noticed that sticking with the old concepts was taking so long to develop the frontend that it was killing my interest in the project's infrastructure, cloud, and backend purposes.
The project was planned to be done in PHP and HTML, because the initial developer had strong skills in these languages. However, when I had to take over, my learning curve was not as fast as expected and my focus was totally different. Even more, after a month I started to get interested in AI and was craving for an opportunity to use it. This ended up becoming a draining obstacle and motivation killer.
The project consists of two steps:
-** Frontend:** that will interact with the user.
-Backend: store the data and keep it for future analysis and business input.
As a Data Engineer, I had a pretty clear idea of how to create the database efficiently and scalably. Although I didn't know how to combine the backend logic with the frontend UI. Then, after some research and exchanges, I came across Anthropic Courses courses from Anthropic to use Claude better and to its full potential.
For this step of the project I specifically did: Claude 101 and Claude Code in Action.
With Claude 101, I got to know the flaws of the application better and how to logically express my questions to get results aligned with the project. After that, I created a project in Claude and combined all the frontend mockups with my database structure, together with short project documentation, goals, my current Git setup, and which skills I was interested in learning. Using this information, Claude provided a great combination of courses, sources, and steps within my project to develop something more aligned with my motivation and skills.
Claude Code in Action was a more specific course. It explained the steps of Claude Code, which I had already used for some projects but hadn't unlocked its full potential on my own. For me, the dealbreaker was understanding the use of hooks and a small introduction to MCPs. I liked it because I could understand how the AI accesses my computer, and the issues regarding data sharing and how to avoid them when integrating the tool with my project folder.
With this post, I just want to share that once you get to know how to use AI and understand how to implement it in your projects, your work can really improve. That said, it doesn't replace the fundamentals — critical thinking, validating results, understanding programming languages, and reading documentation still matter just as much. AI works best when you bring those skills with you.
This reminded me of something a professor once told me during my master's thesis: "The future won't only be about knowing the tools and technology, but understanding the business to know HOW to apply them in the best way possible." It took me a long time to fully understand this, but after this experience, I can safely say I'm finally starting to believe it.
Top comments (0)