Back to Singular
AI as augmentation

You don't need an AI strategy to start. You need one annoying process and a week.

Nina BrenesNina Brenes··6 min read·Updated Aug 2026
You don't need an AI strategy to start. You need one annoying process and a week.
Key takeaways
  • Costa Rica has the highest professional AI-skills penetration in Latin America, more than double the second country (CEPAL, 2025). The skill is not the problem.
  • Do not start with a tool. Start with one recurring process that already annoys you.
  • The first month is small on purpose: pick the process, write down what good means, run it with a check, decide if it stays.
  • The knowledge that makes your work good lives in your head. The move is writing it down as steps AI can follow, so the method leaves your head without leaving your hands.
  • Care about privacy is not bureaucracy. Do not feed sensitive client data into random free tools, ever.

Picture the tab you have had open for a month. "AI strategy for my practice," a document that never gets past three bullet points because every time you sit down with it, it demands a bigger answer than you have. That tab is the problem. Not your discipline, not your talent, but the belief that a plan has to come first. To start using AI in your own work here in Costa Rica, whether you are a doctor, a consultant, a coach, an educator, or an agency of one, you do not need a strategy, a committee, or a full overhaul. You need one recurring process that already annoys you and a single week to run it with AI while you check the output. That is the entire first step. Every post telling you to map the whole shift before you touch anything is, quietly, the reason you have not started. You have been handed the order backwards.

Costa Rica already has the skill. What's missing is the system.

Start with the good part, because it is genuinely good: Costa Rica has the highest professional AI-skills penetration in Latin America, more than double the second-ranked country (CEPAL, 2025). Read that as permission. The bottleneck in your work is almost certainly not that you cannot learn the tools. It is that your expertise is pouring into a way of working nobody ever redesigned for AI. You do brilliant work inside a method that lives only in your head and repeats itself by hand every week. The knowledge that makes your work good is real, and it is yours. The move is not to hire your way out of anything. It is to get that method out of your head and into steps AI can follow.

A row of process steps with one step highlighted as the place to start
Start where the friction lives, not where the newest tool points.

Start with a process, not a tool

The most common mistake is starting from the tool: "let me try this ChatGPT, this agent, this platform." It feels like progress because it is fun and it is shiny, but it quietly asks the wrong question first, what can this do, instead of the right one, what do I hate doing. Flip it. Pick a process you run often and quietly dread: the weekly report, the same kind of reply you rewrite from scratch, the notes you tidy after every session. Just one. The one that makes you sigh before you open it. Say you are a consultant who writes a similar client update every week. That is the kind of process to point at first.

  • How often do you do it? The more often it repeats, the more a week of setup pays you back. A monthly task is a weak first bet.
  • Can you describe what done well looks like, in words? If you can say it out loud, AI can learn it. If you cannot, that is a sign the judgment is not ready to hand over yet.
  • What breaks if it is wrong, and who finds out first? If the answer is a client, before you do, it stays human-checked. Always.

What the first month looks like

In practice it is almost boringly concrete. Week one, you pick the process and write down, before anything else, how you do it well, in plain steps, and what "good enough" actually means. That written version is the whole trick. It is your method, out of your head and onto the page, where AI can finally follow it. Week two, you run it through AI and compare against your bar. Week three, you adjust the steps. Week four, you decide whether it earns a permanent place in your week. That is it. Small as it is, you just turned scattered effort into a repeatable process, while everyone else is still editing the strategy deck.

One warning

Do not drop sensitive client data into random free tools to save a step. Caring about privacy is not bureaucracy. It is what lets you sleep the night after you have automated something that touches real people. Decide from day one which tools you trust with client information and which you do not, before the first scare, because the scare is exactly the thing this care exists to prevent.

FAQ

How can an independent professional in Costa Rica start using AI?

By picking one recurring, high-friction process, writing down how you do it well, and running it through AI for a month while you check the output, before thinking about any grand strategy. Start small and provable.

How much does it cost to start using AI in my practice?

Less than the strategy decks suggest. A first process can run on tools you likely already pay for. The real cost is the time to write down what good looks like and to keep yourself in the loop as the check. The tools are rarely the expensive part.

What do I actually need to start?

Not a big budget or technical background. You need one clear process, a written definition of good, yourself as the person who owns the result, and basic care about what data you feed into which tool. The tools are the easy part.

Where should I start with AI?

With the single process that costs you the most time and patience, not with the flashiest tool. One process, one month, and you as the owner of whether it stays.

About the author
Nina Brenes

Nina Brenes

AI partner for thoughtful experts · Founder of The Harmony Labs

Nina helps thoughtful experts turn what's in their head into AI-powered work that still sounds like them. She reads the research, tests the tools, and sits in the conferences so the people she works with get the short version: what is real, what to try this week, and what to ignore. Her focus is practical judgment, not hype.

Connect on LinkedIn

Stop experimenting.
Start operating.

Tell us where the momentum stalled and we'll scope the engagement.