Vaca Muerta Produces Data: What the Industry Does With It

Vaca Muerta has every ingredient of an industrial-scale AI experiment: thousands of wells, dozens of sensors per rig, instrumented transport networks, and field operations generating operational data in real time, around the clock. Even so, more and more Oil & Gas operators in Argentina are asking themselves an uncomfortable question behind closed doors: how much of all that is actually used to make better decisions?

In most cases, less than it should be.

The technology investment is there. Digital infrastructure in Argentina's upstream grew more in the last three years than in the entire previous decade. The issue is that capturing data and using it to make real-time operational decisions are two different capabilities, and the industry grew fast in the first without solving the second.

Unused Data Is Just a Storage Cost

Vaca Muerta's capex in 2026 is breaking records. In May, YPF applied to join the RIGI with its USD 25 billion LLL Oil project. That same month, Vista raised its investment plan for 2026-2028 to USD 5.6 billion.

At that scale, the gap between using and wasting operational data is measured in points of cost per barrel, and it stops being a matter of marginal efficiency.

The Question First, Then the AI

Operators getting this right share a pattern. They started by defining the operational question they needed to answer: which pump should be serviced before the next maintenance window, which well is underperforming, where the network is losing pressure. Then they identified the data needed to answer it and how to make sure it arrived clean and on time.

AI came last, as the logical close of a workflow that already worked.

Designing for Operator Trust

That sequence requires systems design capability: knowing what to integrate, how to structure it, and how to build the decision layer on top so field operators trust the output.

Trust is the most underestimated part. The system a team abandons in its first week because it throws alerts that make no sense is the most expensive AI project there is, even if the pilot was cheap.

How We Work

At Santex, we bring that approach to energy projects where we integrate data, systems, and processes to improve visibility and decision-making in Oil & Gas operations. We start with the decision the client needs to make, and from there our data engineering team defines what to integrate and how.

Outside the upstream, the same approach guided our work with an Argentine gas distribution company serving more than 87,000 customers a month. We implemented conversational AI for its customer support, with a projected annualized IRR of 71.64% and an estimated payback period of 2.25 years.

The Data Is Already There

The opportunity is open. The industry has the data. What's missing is turning it into a real operational capability, the kind that keeps being used after the demo ends and the dashboard is no longer new.

If you want to review which decisions in your operation could be backed by the data you already generate, let's talk.