HI! I AM
Full-stack developer focused on shipping fast in the AI era.
I spent 5 years in tech consulting, which took me through all kinds of stacks, with and without AI.
Come check out my projects and let's talk.
Website and management platform for Artigas Imobiliária, a real estate agency: from the property catalog to marketing, all in one dashboard.
Personal finance app built on Open Finance: it connects your accounts, organizes your spending on its own and turns budgeting into a game.
My favorite artist is Mac Miller. I love him so much that I have a tattoo of him on my arm. I keep wondering what the songs he never got to release would sound like. RIP Mac
I don't even like to think about how many hours and how much rage I've spent on these two games, but by far, my favorite games are FIFA and Counter-Strike. I was once in the Elite division in FIFA (the highest tier) and I once played against MIBR (and we won) in CS. Today I barely play anymore, but I like to believe that both helped me in ways I can't quite put into words, especially in my career.
THOUGHTS ON THE WORK
Current points of view, but it's always a roller coaster. Not rules, just what has worked for me.
05 takes, best read slowly
I started coding without any AI, started college in 2018. The first time I had to use jQuery in my life I almost gave up on my entire career. I never understood it, and to this day I still don't. I got lucky that syntax and code "got cheap", but whenever I needed to use something I didn't know how to use, I'd figure it out. I'm not a designer and I think this site looks great, I figured it out. I've read books and deeply studied ML concepts, how gradient descent, RNNs and backpropagation work, but I DO NOT know how to train or explain an ML model from scratch off the top of my head, I just, always "figured it out"
When I started using the new AI tools, I used to delegate the research to them as well, and the VAST MAJORITY of the time I did that, my results were worse. Spending a LOT of time laying the groundwork is essential; after that, it feels like the tools "incorporate" your intent, especially on the visual side, but also technically. GOOGLE is still the first step, regardless of which tool I use next.
Obviously, especially working right at the intersection of business & work, it is extremely common and healthy for products to change direction, shape, concept, and purpose as they are built. However, scopes that are too broad, subjective, or formless tend to produce a product that tries to do everything and usually does everything poorly. That is why it is essential for the scope to be defined and refined before starting to build, and constantly revisited throughout development.
It's important from the start to know at least roughly what you're aiming to build, but seeing the pieces like Legos and SEEING THE ORDER OF THOSE LEGOS is even more important. I've done several projects that failed precisely because I didn't take care of and pay attention to every single Lego part. Sometimes it was a good project, but the design was already so messy and bad that a refactor would be hell and nobody could use the product. I've built gorgeous screens with broken or half-baked features. It will always be essential to make the pieces fit together well and IN ORDER.
I've suffered a lot getting attached to code I spent days writing. You spend ages building an abstraction, a well-crafted page, a feature that (often) nobody even asked for and you just thought was cool, then the client changes their mind or the business simplifies and it all loses meaning. That's almost like grieving, and it's still hard not to go through it. I don't think I've found the balance for this dilemma yet. Trying to explain absolutely everything without a demo or a visual mockup can get all your ideas rejected, but jumping straight into building them and working 14-hour days every time will just make you suffer. I still don't really know what to do about this...