One year building KaiNext. Honestly.

Since childhood I wanted my own company. I didn't know what it would be. I didn't know how. I just knew I was passionate about two things — technology and business — and that at some point they would have to come together.

What I didn't know was when or how that would materialize.

Since I was 22 I had never gone a single day without a job. I entered the tech world in Chile in 2017 and never stopped. Offers always came, there were always options. When I was fired on April 17, 2025, my ego was at its peak. My first reaction wasn't fear. It was almost indifference: I'll find a job in two weeks, I thought. Three tops.

But that came later. On April 17 what I had was time, one client, and for the first time in years, no excuse not to do what I'd always wanted.

I called my father

Weeks before the firing I had closed a deal with Patricio, a product distributor who sells from the Maule Region across all of Chile, supplying minimarkets and supermarkets. His business moved millions per month managed between an Excel spreadsheet with a structure only they understood and an app called Mi Negocio that required creating a user for each salesperson and updating stock manually in each profile. He had been insisting for a while that I build something for him.

I had the deal. I had the idea. What I didn't have was the company to invoice it.

The first thing I did was call my father.

Not to ask for money. Not to ask for anything. Just to tell him. He was the only person in the world I could say "I was fired" to without feeling fear or shame. The only person I knew wouldn't judge me, who would listen and tell me to keep going.

And he did.

The days that followed were intense on parallel tracks. I was moving full speed on the project while at the same time defining the company: name, concept, business type, services — everything you need to incorporate a company in Chile.

On April 21 my father came to my apartment at night. Certified accountant, commercial engineer, MBA, more than thirty years in accounting and finance. We sat down, he explained that the most efficient structure was a limited liability professional services company between the two of us, with VAT-exempt invoices, and that same night we incorporated KaiNext Solutions Limitada. One of my bedrooms was registered as the company's office.

Fired on Thursday. Company incorporated the following Monday.

My father has been the person who has supported me the most throughout this entire process, and throughout my life. That is also part of KaiNext's story.

The client I almost walked away from

The project was technically clear. Centralized inventory, sales, commissions, electronic invoicing, branch reports. I built disrover.com, a platform from which they now manage everything. They eliminated Excel. They left Mi Negocio. If they wanted, they could scale toward ecommerce with the same products they sell in physical stores.

The hard part wasn't the technical work. It was learning to work together.

Patricio built his business through sheer effort, hard work, focus, and consistency — waking up early every day, building relationships with his clients and maintaining them over time. That's what built his business. But his world and mine are different. The way he communicated what he needed and the way I understood requirements didn't align naturally. There were voice messages at six in the morning. Tangled conversations. A mid-project negotiation I didn't see coming.

There was a moment when I was ready to tell him I was done.

I didn't. Both of us had to make the effort. He had to adapt to my timelines and my way of working. I had to learn to read him, to understand what he needed behind what he was saying, to communicate in his language. It was uncomfortable while it lasted. In the end we made it: him with a system that transformed the operation of his business, me with my first project delivered and paid.

The system runs in production today, twenty-four hours a day, seven days a week.

The arc with Patricio didn't end there. In February 2026 someone notified me that disrover.com was redirecting to another page. I did the forensic analysis, identified how they had gotten in, implemented the fixes. In March it happened again, this time with the attacker still active, looking for new entry points. I migrated the entire infrastructure to a clean environment and left a sentinel monitoring it.

Since then, no incidents.

Patricio had chosen not to hire active paid support. That's his decision and I respect it. What I learned: when a client chooses not to contract protection, the problem doesn't disappear — it just changes who it falls on when it happens.

50,000 searches and a product I shut down

In May 2025, while Disrover was moving forward, I launched micodigopostal.fun.

The problem was simple: in Chile there was no fast, user-friendly way to look up your postal code. I built something modern — a clean form with browser-based location access for instant lookup, connected to an API that scraped the official Correos de Chile form, saved the result in our database, and returned it to the user.

I posted on LinkedIn. It exploded.

2,832 reactions. 146 comments. 135 reposts. I added it to KaiNext's portfolio, kept it running for several months, and ended up with nearly 50,000 postal codes registered by users themselves.

And then I took it down.

My reasoning was that it had no clear business model and it wasn't worth continuing to pay infrastructure for something going nowhere.

As I write this I realize I was probably wrong. 50,000 real searches, organic traction without a peso invested in marketing, a concrete problem solved better than the official alternative. Maybe at some point I'll bring it back, with more experience and a clear business model to take advantage of it.

What I learned: LinkedIn hype is not business validation. But 50,000 users doing something real in your product is. Sometimes the difference between a dead product and a live one is knowing how to read which one you have.

The second client came without marketing

The second client didn't come from LinkedIn or Google. It came from something you can't buy.

Andro Ostoic and I met at BancoEstado, where we were both software architects. We had worked together inside the bank and on external freelance projects. We had built MiBE+ together, a mobile app for the bank. We knew each other well — not just technically, but in the way we worked, solved problems, communicated.

When he found out I had incorporated KaiNext, he told me he was in exactly the same situation: working at the bank while building his own company, ABCS PRO STUDIO. He had landed the Art Santiago project, a mobile app. He handled everything: backend, web frontend, infrastructure, deployments. But he needed hands on mobile and thought of me.

Not a freelancer. Me.

The product went live on App Store and Play Store in August 2025. KaiNext's second client.

What stayed with me from that: before I had a company, I was already building the client base without knowing it. Every project done well, every colleague you didn't let down, every problem you solved when you could have walked away — all of that is reputation. And reputation is the asset that doesn't appear on any balance sheet but is the first one to generate business.

The job that lets me build without desperation

With Disrover delivered and Art Santiago underway, my unemployment insurance had an expiration date. I needed a job.

And that's when I truly collided with reality.

Technologically I've always been on the edge — always reading, always testing, always improving my workflows. That wasn't the problem. The problem was the market. AI advances had raised the bar. Companies were doing mass layoffs. There were more good engineers looking for work than available positions. To land the salaries I wanted without lowering my standard of living, I needed an English level I still didn't have. I needed my digital presence in order: LinkedIn, GitHub, CV. I had to prepare to compete for real.

I applied to several jobs. Some never called me. In others I got through interviews and didn't make it. Some processes were disrespectful of the candidate's time — technical tests, interviews, silence. That also teaches you: to value the spaces where they treat you as a professional from the first minute.

There was a moment when I thought: I'm scared and I'm willing to lower my standards to have a stable income soon.

That moment passed.

Thoughtworks was my last shot.

Karina Palacios, a friend I deeply admire both for her career and for who she is as a person, wrote to tell me her company was looking for people. She sent me the roles. I agreed to apply. She recommended me and the process started.

I put absolutely everything into it. The process took about a month. I was scared, more than I like to admit. I knew that if I didn't make it I would have to rethink everything.

I made it. On August 4, 2025 I joined as Senior Consultant. Two days later I was on the LATAM Airlines infrastructure team.

Since then I've met engineers who have been searching for months, some more than a year, for formal employment. I was fortunate and I know it. Today Thoughtworks feels like home. I feel like I found my place, and I want to keep building there.

KaiNext didn't stop. It kept going in the margins of the day.

What I learn in the corporate and consulting world I apply to KaiNext. What I build at KaiNext makes me better at Thoughtworks. It's not sacrifice or postponement — it's a real, conscious cycle where both sides benefit.

The direct consequence is that KaiNext grows more slowly. I know it and I accept it. The goal is not to jump to KaiNext full time. The goal is for KaiNext to become real additional income that at some point equals or exceeds my salary, while keeping Thoughtworks. The scenario of 100% dedication to KaiNext exists, but it requires KaiNext to justify it with numbers, not enthusiasm.

Tools I built for myself

From real cases I was solving throughout the year, I built AI tools. First they served me. Then I opened them on kainext.cl because if I had that problem, others did too — and that way I could scale without creating a new project from scratch each time.

In that process I integrated MercadoPago as a payment method for websites. I validated it the only way that gives you real certainty: I used my own card and paid one thousand pesos for one credit. That payment was KaiNext's first real client in that model.

The first external payment hasn't arrived yet. The tools today have four registered users. Two are me — an admin account and a regular account. The other two are trusted people who helped me test them.

It's honest to say so.

What I take from the process isn't the numbers — it's the learning. Building something to solve your own problem and then opening it for others is the most honest way to do product. I knew exactly what it solved and what it didn't, because I used it in real conditions before publishing it.

The path to the right tool

My sister Silvana is studying law and works at a law firm. In March 2026 she told me she had an enormous task ahead: manually reviewing and cross-referencing hundreds of legal documents spread across dozens of files. They had hired external help to do it. Manual work, row by row, document by document.

I heard that at a family birthday and that night I already had the architecture mapped in my head. By Sunday, March 15, I had a prototype deployed. Forty-eight hours from the birthday to the deploy.

The problem was that I had built the solution to the problem I understood between empanadas and drinks, not the real problem she faced every day at her desk.

I asked her to explain everything in detail. She sent me via WhatsApp a fifteen-minute voice note, a two-minute one, and a seven-minute video. Twenty-four minutes of content. But processing twenty-four minutes of audio by hand — pausing, noting, rewinding, listening again — was exactly the kind of slow process I wanted to eliminate.

That's when I asked myself: why am I going to transcribe this by hand if I can build something that does it in thirty seconds?

That night I built an AI audio transcription tool. At 2:21 AM on March 30 I had the MVP validated with Silvana's real audio files. The fifteen-minute note became a four-page PDF with every detail I needed. Not a single word was lost.

The tool I wanted to build for Silvana was one for intelligence on legal documents. I got as far as having it in the KaiNext dashboard. Architecture defined, data model, complete implementation plan.

On April 22, 2026, I cancelled it.

Silvana had resolved the problem on her own. Without a real user with a problem they couldn't solve alone, there's no product — there's a technical exercise. Continuing would have been continuing out of inertia.

What I take from it: sometimes the path to the right tool forces you to build another one first. And sometimes that intermediate tool turns out to be more useful than the one you had in mind. The transcription tool is now available on kainext.cl, because if I needed it, others do too.

April 21, one year later

On April 21, 2026, my father wrote in the family group chat. He had made a decision: he wanted to dedicate himself fully to accounting, tax, and financial consulting — find his own clients, build something of his own. He was also finishing a Master's in Tax Law. The decision was made.

My first reaction was pragmatic: I told him it was risky, that the market was difficult for everyone.

He answered that he was going all in.

I told him he had my full support.

It wasn't just his story. It was exactly mine, one year earlier. The same point of no return. I had also let go of a job without certainty of what was coming. And what came was KaiNext.

That same week I asked him if he wanted to appear in KaiNext. Not as background support, as he had been until now, but in front — as a visible part of the company.

I wrote to him: "If you're interested, I could now orient KaiNext toward offering more integrated services — optimization and business audits at both the tech and financial level."

His response came within minutes: "That would be amazing!"

Two messages. No meeting, no presentation, no deck.

For the first time the model had the two pieces it had always been missing together: technology and finance, under the same brand, built by the two people who truly understand them.

On April 21, 2025, we incorporated KaiNext together. On April 21, 2026, my father decided that he too was going to build something of his own.

KaiNext always had both parts. It just took both of us being ready to see it at the same time.

The second year starts now

KaiNext is one year old. Incorporated April 21, 2025. Not yet profitable.

But this year truly existed. A system running in production for a real company. An app on App Store and Play Store. Tools I use myself before opening them to others. A product that generated real traction and taught what it had to teach. Clients that arrived without a peso invested in marketing, solely from having done good work before having a company.

And everything I learned this year.

How to negotiate with a client who sends voice messages at six in the morning. How to do a forensic analysis after a hack with the attacker still active. How to build a tool for the wrong problem, realize it, and try again. How to kill a project you've already invested time in without letting that paralyze you. How to compete in the hardest job market I've ever faced in my life. How to value a job in a way I'd never felt before.

None of that was in the plan. Because there was no plan.

The second year starts now. With Thoughtworks as my professional home. With KaiNext as my company. With my father as my partner. No guarantees, no clients in the pipeline yet, no profitability.

But with a full year of learning that doesn't disappear. And with the certainty that the only way to truly learn this was to live it.

I started this year being fired. I'm ending it building a firm with my father.