A couple of months ago I launched Código Chigüiro from Villavicencio — not from Bogotá. An AI consultancy serving companies in the Colombian Llanos.
Before building the first solution, I spent weeks talking to people who are already in the field: owners of oil & gas contractors, SMB managers, recruiters, cattle ranch owners, officials from small municipal governments.
The question was simple: what's going on with this AI wave for them? The answer was different from what I expected. Here are 3 things I learned — not in theory, but by listening.
Business owners in the Llanos don't run from AI. They run from the fluff.
When I sat down with the owner of an oil contractor in Meta, I expected to hear «that's not for us» or «too complicated». Instead he said: «Send me whatever you want, just not a presentation with the words digital transformation on it».
That line stuck with me. Because I heard an almost identical version 4 more times — from SMB managers in Villavicencio, from recruiters in Bogotá, from a mayor of a small town.
The problem isn't the technology. It's how it's been sold to them. For years they were offered «comprehensive AI solutions», «intelligent platforms», «cognitive ecosystems». Nice words, zero measurable impact. They spent money. They saw no results. Now they distrust it — rightly so.
What works: close the slide deck, open the laptop, show a real process. Here's a folder with 300 invoices. This AI reads them in 2 minutes. Now your team can review the 5 weird cases instead of loading all 300. That convinces. Diagrams don't.
The problem isn't not knowing AI. It's not knowing what to automate first.
Second question I ask in every conversation: «If you could automate one process in your company, which one would it be?». 9 out of 10 times: long silence, then «hm, I'd have to think about it».
It's not that they don't know. It's that they know too much. They have 50 processes they'd love to automate and they can't prioritize. It's analysis paralysis — and it makes sense, because prioritizing wrong costs money.
The trap for founders with budget is wanting to automate everything at once. They hire someone, set up a committee, run a 4-month assessment, and end up implementing nothing. I've seen this 3 times in the last month.
What works: pick ONE process. Just one. Measurable, short, repetitive. Automate it in 2 weeks. Measure the real savings. Show the numbers. Then decide the next one. Automate one, prove it, iterate.
«You don't know my sector» is the most common objection. That's why I work by sectors.
Third thing I learned: the #1 objection in LATAM isn't price. It's «you don't know my sector». And it's a legitimate objection. A generic chatbot doesn't understand that when an oil contractor says «work front», it means something different than when a retail SMB says «front» — they might not even use that word.
An AI model that works in banking doesn't work in cattle ranching, because the vocabulary, the regulations, the concerns and the daily dynamics are radically different.
That's why Código Chigüiro works with 6 specific sectors in the Llanos: oil & gas, local SMBs, HR, cattle ranching, municipal government, and healthcare. We don't offer «AI for companies». We offer AI for oil contractors in Meta. AI for SMB owners in Villavicencio. AI for cattle ranches in Casanare.
The cost of that decision: we can't serve every sector. The benefit: when we walk into an oil contractor, we already know what the HSE report is called, what Ecopetrol asks for, what format they use. The sales conversation is 5x faster and the client trusts us.
The reflection
AI isn't the future of the Llanos. It's the present of the Llanos that's still not being told.
There are already oil contractors in Meta using OCR to extract data from HSE reports. There are SMBs in Villavicencio with WhatsApp chatbots that answer 80% of the queries. There are recruiters in Yopal filtering 200 resumes in 3 minutes. It's happening. It's just that nobody's telling the story — because the AI narrative is dominated by the big ones, the ones from Bogotá, the ones from Silicon Valley.
Part of Código Chigüiro's job — aside from building solutions — is to tell those stories. To make visible what's already working in the Llanos. So the next business owner who thinks «that's not for us» has at least one case to look at.
Next chapter
If you're in any of the situations I described — distrusting the fluff, paralyzed by not knowing where to start, or convinced your sector is too unique — write to me. I'll listen first, before offering you anything.
The next article in this series («Notes from the Llanos #02») comes out in 2 weeks. I'll tell what I'm learning as I make the first real sale. If you want to follow it, I publish on LinkedIn first.