KLIMMER KLIMMER
Ascent Logbook · Article 06 · AI

AI doesn't choose the summit

Artificial intelligence as a tool, with the course in human hands

By Cristóbal Almeida, Managing Partner, Klimmer Consulting7 min readSeries: Lessons for scaling

Imagine telling an artificial intelligence: “I'm the new CFO. Analyze all the company's financial information and tell me, with red, yellow and green traffic lights, what I need to focus on for the next three months, with an action plan”. And in seconds it hands you a dashboard with priorities, risks and next steps.

That's no longer science fiction. It's possible today, as long as the AI is connected to reliable, integrated, up-to-date information. An analysis that would take a consultant weeks of gathering data from isolated systems can be done in the time it takes to pour yourself a coffee.

A tool, not a business model

It's worth putting AI in its place. For the vast majority of companies it's not a new line of business, but a tool for their business. It doesn't come to run the company without people; it comes to take mechanical work off their hands and give them better information to decide.

But careful: being a tool doesn't make it optional. Companies that adopt it with structure will win market share from those that don't. The advantage won't be having AI, but knowing how to use it on top of an organized operation.

Artificial intelligence can take you along the best path. Deciding which summit you're heading for is still human work.

Only as good as your data

Here's the part almost nobody tells. An AI connected to fifty spreadsheets that contradict each other doesn't produce intelligence: it produces confusion faster. If the information is fragmented, incomplete or arrives a month late, the AI inherits all those problems.

That's why the sequence matters:

  • Processes: understand how information flows and organize it.
  • ERP: a single central map where that information lives integrated and in real time.
  • AI: the layer that analyzes, alerts, automates and recommends on top of reliable data.

Skipping the first steps to go straight to AI is like hiring the best guide in the world and giving them the wrong map: they'll take you very fast and very confidently… to the wrong place. AI amplifies the quality of your information; if the information is bad, you just get it wrong faster.

A professional working with data on his computer
AI multiplies what you already have: if your data is solid, it multiplies clarity.

Knowing how to ask is the new skill

AI still requires your intelligence to get the best out of it. It doesn't think for you: it answers what you ask. A vague question gets a vague answer. A question with context, a role, a time horizon and a clear format gets a useful answer.

“Analyze my sales as if you were my competitor and tell me where I'm vulnerable.” “Detect which customers have changed their buying pattern in the last 90 days.” The limit is no longer technology; it's the imagination and judgment of whoever asks.

Autopilot takes you wherever you tell it

It's tempting to imagine an AI that doesn't just recommend but executes: that collects, buys, adjusts prices and optimizes everything to maximize profitability. Some of that will come. But even the best autopilot needs someone to tell it where to go and what limits to respect.

A company exists for something more than optimizing a number. Its purpose, its values, the way it treats its people and its surroundings, the lines it isn't willing to cross: that can't be delegated. AI knows the route; you choose the summit.

A person on a mountaintop facing the light
Direction, purpose and limits are still human decisions.

Where it does create value today

  • Financial and operational analysis in seconds, with clear priorities.
  • Early warnings: collections at risk, stalled inventory, eroding margins.
  • Automation of repetitive tasks: turning an email into an order, reconciling, classifying documents.
  • Summaries and reports for leadership, without waiting for the month-end close.
  • Sharper demand and cash-flow forecasts.

Four questions before investing in AI

  1. Does my information live in one place or scattered across files?
  2. What specific decision do I want to make better, or what task do I want to stop doing by hand?
  3. Who on my team will ask the AI the questions and validate its answers?
  4. What limits (ethical, regulatory, brand) must it always respect?

The KLIMMER view

Our tagline says it in order: Processes | ERP | AI. At KLIMMER we help companies apply artificial intelligence where it creates savings and better decisions, and prepare the data and process foundation that makes it work. No magic promises, and without losing sight of the essential: technology speeds up the climb, but you set the direction.

Share

Where does your company stand today?

Start with a diagnostic. We review your checkpoints, locate the leaks and define the route to your next level: processes, ERP and AI, in that order.

Keep reading

More articles in the series

Process, technology and operational AI consulting. Monterrey, N.L., Mexico.

Message us on WhatsApp