Author's note (May 2026): This article was originally written in my personal notes in September 2011, in Guanare, Venezuela. I'm sharing it publicly for the first time here, with minor clarity edits.

The Moment That Changed Everything: Why I Decided to Build a Sign Language Translator

The afternoon walk home

It was an ordinary afternoon in Guanare. As usual, I was walking home from school — a 60-to-90-minute walk I made every day. Halfway there, I ran into a friend I hadn't seen in a few days. She was heading to a self-defense class at a gym along my route, so we kept walking together.

At some point, we crossed paths with her cousin. He was her age. They greeted each other, then both looked at me. My friend began moving her hands and making facial expressions — sign language. She was telling him who I was and that I was walking with her. Her cousin smiled and signed back. I stood there watching, completely locked out of that conversation.

After a moment, he said goodbye and went his way. She and I kept walking.

I was immediately curious. "What did he say?" Her answer was brief: "If you want to know, figure it out yourself."

Something struck me clearly in that moment: without understanding sign language, I was the one who was limited. Not her deaf cousin. Me. In a conversation as trivial as introducing an acquaintance, I was the excluded one. And that felt completely unacceptable.

What I searched for that week

My curiosity didn't fade the next day. I saw my friend again a few days later, and her first question was whether I'd figured out what they said. I had to admit I hadn't — but that I was looking.

The closest thing I could find to gesture detection by computer in 2011 was the Xbox Kinect — and a community of enthusiasts hacking and modding it for new purposes. But everything was in English, and the hardware was hard to get in Venezuela.

I also read forum posts about translation systems using sensor gloves or wristbands. But I sensed those weren't enough: sign language isn't just hand movements. It's complete sentences, facial expressions, body positioning. It's a complex visual language, and reducing it to finger sensors felt inadequate.

That's when I understood: something would need to be built from scratch.

The idea that took shape

The connection was natural: if the Kinect could detect body movements to control a video game, why couldn't it detect sign language gestures and translate them? If it was possible for gaming, it had to be possible for communication — something infinitely more important.

I had no idea how to program it. But I knew that computer vision could detect and interpret movement. And that was the core of what I needed to build.

The context: Venezuela in 2011

What moved me wasn't just technological curiosity. It was also the context surrounding me.

Venezuela in 2011 was politically and socially difficult, especially for someone with limited resources and an idea that was both necessary and technically complex. The deaf community had no way to participate in public life or spread messages beyond their own sign-language circle.

Near my home, in the 23 de Enero neighborhood, a pilot school project for people with disabilities had been started. But like so many government projects of that era, it stalled — no specialized teachers, no resources to complete the infrastructure, no follow-through. What could have been a good starting point to connect with the deaf community never materialized.

Realizing there was no functional digital tool in Spanish to translate sign language — in a country where inclusion projects were born and died incomplete — was the definitive trigger.

First steps

I tried to formalize TALS as a thesis project for my Computer Engineering degree at university. I went in with the idea, with enthusiasm, thinking that academic backing could accelerate the project. The answer was discouraging: if I developed it under that structure, with institutional supervision and support, any rights to the project would pass to the university.

That was the first inflection point. It wasn't arrogance that made me decline — it was the conviction that what I wanted to build was something important, and I wasn't willing to give it away. So I kept researching on my own.

I remembered that some time earlier I'd applied to Venezuela's Fundación Ideas without making it past the pre-selection stage. This time I had no funding and no mentor. I had a basic laptop, limited internet access, and the conviction that if I could get a Kinect, I could build a first prototype and prove — to my friend, and to myself — that the idea was viable.

The commitment

It wasn't an easy decision. Studying Computer Engineering while also completing my Animal Production Engineering degree was already a heavy load. Working to pay for my studies left very few hours each day for the project. And the country was deteriorating: currency devaluations, rising insecurity, political uncertainty.

But I told myself something simple: if I want this to exist, I have to build it. No one else will do it for me.

The tension was real: "this is impossible for one person alone" versus "if not me, then who?" In the end, the second phrase won.

Without being able to acquire the Microsoft Kinect at that point due to lack of resources, I focused on documenting the idea and continuing to research. No money. No team. No idea what an "entrepreneurial ecosystem" was. Only the certainty that the idea was worth the effort, and that when the day came I had the resources to prototype, I would be ready.

That certainty was the first brick of what, six years later, would become TALS — a registered company in Chile, a CORFO Seed Capital Prize worth 33 million Chilean pesos, and a stand at Computex Taipei representing the country that took us in.

It all started one afternoon on a walk home, when a deaf cousin laughed at something I couldn't understand.