who is Gabriel?

I was born 29 years ago, on January 27, 1997, in Porto Velho, in the Brazilian Amazon. The family reputation came early: impossibly stubborn. Conviction and curiosity showed up from the start. I would take apart, break and rebuild the toys and objects around the house from imagination alone, and everything sparked an endless curiosity in me, especially about how the world actually worked.

When I was four, my father, a sales rep for BIC, would sometimes bring home old display cases from past collections, pens of every kind. They grew my collection and planted a thought: could I make some money selling these pens, enough to buy a magazine or a toy I wanted?

My mother carried me along, whenever I was not in school, to her house calls. She sold clothes door to door. In Brazil we call it being a “sacoleira”, because you haul a giant bag of clothes in every style straight into your clients’ living rooms. I sat there listening to the sales arguments. My father and my mother were exceptional sellers: they did what we now call rapport without ever having studied it. It was pure instinct.

And me? Porto Velho had no store with the toys I liked, so I asked for money or saved my allowance to spend at the toy store whenever we visited my parents’ family in Campo Grande. If I sold the pens and saved more, that would be extra cash.

In less than a week, my classmates’ lunch money had become part of my savings for the next trip. Not that it was much. Four-year-old Gabriel had no idea what each pen was actually worth.

I remember my first computer to this day. Late 2001: my father brought home an all-white machine, set it up in my parents’ bedroom, an impeccable and far from modest setup for the time. I even loved the noise the internet made so he could start playing.

My father never went to college, but he had a gift for numbers, persuasion and a phenomenal sense of integrity. And there he showed me he was good at technology too. Besides the games, of course. He started playing Counter-Strike. It sounds strange to say, but my father, a sales rep from a humble family who had worked since the age of six, was absurdly good at shooters for his age. His ability to learn far outside his comfort zone always intrigued me.

At six, when he moved on to Chaser, he upgraded, built a new PC, and I inherited the old one. I started playing alongside him. And, strangely, I also started browsing. The idea of research felt like magic: I fell into forums and loved to read. That is how I learned what Cheat Engine was, nothing more than a memory editor, capable of taking my character from normal speed to ultra fast (my father did not approve, let the record show).

That opened very specific doors in my head about what was possible. On a trip to Campo Grande, I watched my older cousins playing Tibia at a LAN house. It was one of the biggest MMORPGs of its time: huge communities, people on TeamSpeak calls for hours, deep role-play. I loved it. And I had an absurd knack for self-guided learning, what I now call “headbutting through”: learning by trial and error, nonstop, until there was nothing left to learn. Or so I thought.

I fell into a forum about alternative game servers, and my life changed. Instead of just playing, I started figuring out how to run my own server and bring my favorite cartoons, animes and mangas into Tibia. I edited the servers’ XML blocks and Lua scripts, drew maps and features, and inched my way toward C++. And I understood that in there, no matter how hard I headbutted, I would never know everything. I would often have to ask the forum for help, trading knowledge, surviving as a community.

Strangely enough, that did not frustrate me. It only pushed me to learn more. I spent entire days on those forums and, at 11 (as crazy as that sounds looking back), I became one of the four directors of XTibia, Brazil’s largest forum for alternative Tibia servers. On alignment calls, an 11-year-old kid gave orders to moderators and community managers, and was respected. Not for his age, not for the title: for the ideas. That only motivated me to keep going.

But I was no longer the kid without responsibilities, and little by little Tibia lost its pull. I stepped down and went on trying to build things that could earn me something. I ran my own Tibia and Ragnarok servers, wrote bots, looting and farming automations... Yeah, not the part I am proudest of. And rich I did not get. If only I could remember where I stashed that bitcoin wallet...

At 12 I started working occasional days at the phone accessories store my parents had opened, To Go. BIC had let all its sales reps go. My father was the last one out, which gave him time to prepare and jump into his own business. I would leave my after-school activities and go straight to the store: every day with at least four hours of work was generously rewarded with 50 reais. Back then, that meant two movie tickets with popcorn and juice included. Or a new game, or in-game cash.

Noticing how often it paid to order things from China for my parents, who struggled with novelty products and suppliers, I started learning how Chinese import markets worked. Only this time I wanted to invest my own money and resell to them. And that is how all the savings from two years of work by young 14-year-old Gabriel ended up in the hands of a scammer on Alibaba. A hard lesson. Which, honestly, I did not quite learn.

My mother wanted me to take school more seriously, and I hated the idea. I ended up forced into tutoring classes at the best tutoring school in the region, even with grades above 9.5 out of 10. I was good at taking tests, at memorizing exactly what was needed right before the exam. I do regret, a little, not being the person who genuinely wanted to learn everything a Brazilian school can offer someone who studies on their own.

Except the tutoring made no sense even to the school’s owner. I knew everything and could explain my reasoning clearly, helping classmates who actually struggled. Seeing that, she invited me to teach other groups. The school had grown too fast and she needed “qualified” staff. I started teaching students two years below me and, before I knew it, I was teaching people from my own class, in the same subjects I was taking.

On top of that, I built websites and set up forums for all kinds of markets, still answered questions about my XTibia materials, and every day I enjoyed solving problems with software a little more. Or frying some board while brute-forcing it...

The money was good, and I threw myself into ambitious projects again. This time I did not want to bring just accessories: I brought novelties like plasma lighters and every crowdfunded gadget I could find. But I bit off more than I could chew and lost most of it in a failed attempt to bring iPhones into Brazil. Another Alibaba scam. Unfortunately, the scammers were good.

At 15 I moved to the countryside of São Paulo for my last year of high school. Our school material in Porto Velho came from Poliedro, and the idea was that, living in the heartland of Brazil’s best college-prep material, I could pass any entrance exam. That was the idea in my parents’ heads. I had no clue what degree I would pursue, or whether I wanted college at all. But their dream was the diploma they never had: both of them had to prove, all the time, that they knew things, because there was no piece of paper attesting to the broad knowledge and skill they carried.

The following years were counter and project: To Go, imports, websites and forums, while I kept postponing the college decision. In 2016 I got into Computer Engineering at FURG, a federal university in southern Brazil. But rewinding a bit: I was never an exemplary student, and at 16 I was diagnosed with ADHD and giftedness. Someone restless, driven, hungry for real-world action, spending five more years inside an institution... did that have any chance of working out? Maybe. In the head of the Gabriel of that time, no.

After one full year, I paused my enrollment and went back to Porto Velho. I was determined to help my parents with the accessories business. They were going through a double bankruptcy: the company’s and their own marriage’s.

I took over the worst operation and turned it around: from a loss of 30 thousand reais a month to revenue above one hundred thousand, running in the black, within six months.

That is when I looked at it all and asked myself: do companies also have problems I can solve? What if it is not just the line of code? The game? Someone struggling to understand a subject?

At 20, I left the businesses I had helped stabilize to build my own, on new ground. Savings gathered, car sold, I threw myself into Campo Grande. Affection for the place, I think: I liked the idea of that city.

I convinced the mall’s board to open a spot for an accessories kiosk, at a time when there were no kiosk vacancies at all, let alone for accessories, which looked like a passing fad. But GEEKS was something else. I wanted to deliver what I could not inside To Go, and the dream was bigger: to become a crowdfunding-driven channel for new technologies, bringing those novelties into Brazilian reality.

It took six years and many battles. Brazil was never an easy place to build a business, and I knew a lot about the front line and little about the back office, the financial management, the contracts.

In those six years I built a brand that the whole city and its neighbors recognized. We became a byword for quality in a commodity market, where any street stall also sold accessories. Our accessories revenue alone matched our electronics sales, and three years in I was closer than ever to the dream.

And technology never left the stage: I automated the inventory-POS flow until stockouts dropped 95%, and started training demand-prediction models to decide purchases before the shelf went empty. The ML that would later become a consulting practice was born inside retail.

I could blame people, circumstances, COVID-19... but in the end, the mistake was mine. I did not prepare, I trusted wrong, I negotiated wrong. By the end of 2022, even with record sales, the operation was unsustainable. I had failed at managing risk. The bankruptcy of my name and my company was certain.

I tried everything to keep it standing.

That is when a side job I already had, consulting for friends’ and family’s companies (never paid; back then I saw it as a stimulating mental challenge that also taught me other industries), turned into something serious.

I set out not only to try to save GEEKS: I used all the machine learning I had built for demand and product prediction, plus everything I knew about business and process, to help other companies and other people. And it worked.

But GEEKS was lost. I could have pivoted the company, but that would not be its idea of itself, nor mine. I know I am drawn to solving chaos, but there were skills still missing, skills that later, because of this scar, got built.

I do not see the bankruptcy as failure. Failure would have been giving up. I pivoted myself, not the company: I went on to help others avoid the management and analysis mistakes I had made. And I started finding joy not in a seven-figure revenue of my own, but in other people’s dreams moving forward.

I already knew language models, machine learning, Python. I never lost the ability to structure systems and solutions. And with GPT-3.5 I saw that this was the future. I dove in headfirst: what I already did in consulting, I could multiply with software. And multiply the companies too.

The journey ran through many projects and companies: predictive market analysis (an Achilles’ heel: a breach in my own system cost me money out of my own pocket, and security was never optional again in anything I built), legal-tech case analysis, AI automations that saved months of work, transforming workflows, security, revenue and scale. I went all in: I built mechanisms for software development at the highest level and also small, focused tools, integrating AI and automation wherever they made sense.

I joined Yorus, an agency focused on performance marketing, in 2025, through the Tauá Resorts project; in early 2026 I stepped up as CTO, to add technology and AI to the marketing practice. That chapter closed in August 2026. For Tauá we built an AI that has passed 293 thousand conversations, resolving 87% of them without human intervention, at a 2.1-second average response time, and that supported e-commerce sales through a black friday that grossed over 40 million reais.

We also built software for real estate developers, agencies and their customer service. I mentored technology directors at legal-tech and hospitality startups. And we delivered a native AI integration inside Salesforce, without AgentForce, with custom apps built for Tauá.

All of it held to the same bar as always: measurable results in production, not prototypes.

Today I run Groundfast, an AI forward deployed engineering practice, and I remain open to the next chaos: a real operation, a messy problem, an AI that has to prove itself in production. That is where I work best: understanding people and business before code, and leaving only when the number shows up.

If your chaos looks like that, talk to me.