Gustavo Sabato, CIO Aeropuertos
In my missive on real-time organizational truth, I argued that enterprise software vendors should apply more data imagination to AI use cases.
And: AI agents can’t be effectively intelligent without relevant real-time context – wherever that context might live.
You don’t build that kind of context overnight – but I’ve been on the lookout for companies that are building agents that combine operational data from so-called systems of record, such as modern ERP, with savvy use of external or unstructured data.
Aeropuertos Argentina – and the launch of the Snow Agent
At SAP Sapphire 2026, I delved into this type of use case with Aeropuertos Argentina. At the time, Aeropuertos was on the verge of launching their award-winning Snow Agent, timed with the pending Argentinian winter. Well, winter has now arrived – and Snow Agent is live. The agent goes live just in time. As per Infobae.com, managing airport winter conditions in Argentina is no joke:
With the support of specialized machinery and a team of 20 people, antifreeze chemicals, such as solid urea and liquid glycol, are applied. These products can remove ice and prevent its formation at temperatures as low as -50°C and -8°C, respectively. This is done while continuously monitoring weather forecasts and real-time weather conditions (temperature, humidity, precipitation, and wind). [Author’s note: these quotes are translated from Spanish by Google Translate].
Enter the Snow Agent:
Within this process, Aeropuertos Argentina developed the SNOW (Smart Network for Operative Winter) solution , internationally awarded by the company SAP, which allows anticipating incidents and automating their management, improving the effectiveness of the operation and strengthening sustainability, since it avoids the waste of urea and glycol.
The potential to blend crucial external/unstructured data with enterprise workflows is where the real agentic action is. The SAP Innovation Awards explains how this works:
Snow Agent addresses the complexity of winter airport operations by orchestrating the full snow and ice management lifecycle. During adverse weather conditions, the solution continuously analyzes real-time meteorological data and AI.
Based on these insights, Snow Agent automatically validates the availability of materials, machinery, and human resources, creates and releases inspection or contingency work orders in SAP, and tracks execution in real time through mobile-enabled processes. By embedding intelligence directly into SAP maintenance and supply chain workflows, Snow Agent ensures faster, safer, and more consistent operational decisions during critical winter scenarios.
But how did we get here?
Many enterprises don’t know where to start with agentic rollouts.
The modernization of applications – not to mention data platforms – is an underrated ingredient to quality AI agents. Heck, it might even be a prerequisite for the kind of agents that deliver value, rather than probabilistic attempts at reliability.
Building AI agents – on a modern ERP foundation
At SAP Sapphire 2026, I went behind the news with Gustavo Sabato, CIO, Aeropuertos Argentina. Sabato’s SAP experience goes back to 1997. That experience matters now: Aeropuertos Argentina must support the safe operations of 35 airports in Argentina, including 90% of the commercial flights, and more than 43 million passengers a year.
The Snow Agent launch doesn’t come out of nowhere. It is built off the foundation of Aeropuertos’ global SAP migration push to S/4HANA – with the plan of operating off of a global business template. Sabato says Argentina moved to S/4HANA in 2023, then Armenia and Ecuador, with more country go-lives on the way. This is now officially a RISE project – but why make these moves? As Sabato told me, ECC was just not the right platform for their AI plans. However, the cloud wasn’t the core motivation:
Instead of thinking if it’s cloud or not cloud, something that for me was more important is: we need to do the global template for the business perspective.
With this foundation in place, forward-thinking projects like Snow Agent are viable. Working with SAP, Sabato and his team built the Snow Agent in 12 weeks. Any reduction in “interventions” on the runway is going to save money, time, chemicals, and passenger inconvenience. Sabato says that with 1.4 million people using these airports 90 days, they would end up with hundreds of weather-related interventions on the runway:
The problem historically is the processes are manual or fragmented, and it’s clear that this affects your operational cost, your administrative cost, and the use of the chemicals that you need to put out to clean the ice and to clean the snow… If you have more interventions, you need to use more more equipment, more people, more money, and affect the environment.
So how can the Snow Agent change this? As Sabato told me, it’s about pulling together disparate information, including sensor data, to generate alerts and track incidents:
[With the data] coming from the sensors, the agent starts taking this information, and then generates alerts, generate work orders, or releases work orders,*** and checks the availability of the truck resources that we need. Also, it helped us to maintain the communication between the control tower and the people that are working.
The wrap – on ERP transformation and agentic AI
If all goes well, what results will Aeropuertos Argentina achieve? When it comes to passenger safety in inclement weather, there is much more to it than ROI. Aeropuertos projects a 45tCO2e reduction in annual carbon footprint (metric tonnes of carbon dioxide equivalent), a 16 percent reduction in direct costs, and a 90 percent reduction in administrative time.
But this isn’t about a single AI agent. The SAP Innovation Award quotes from Aeropuertos pointed towards the winning combination of apps/platform modernization and agent building. As Sabato says:
We began a deep technological transformation with S/4, and that foundation opened the door to new systems, new processes, and a new way of working
CEO Martin Guadix talked about the push from fragmented data to end-to-end traceability:
In a context where operational decisions are time-critical, Snow enables us to turn fragmented data into actionable insight, automates notifications and work orders, and secures full end-to-end traceability to sustain safe and efficient operations under critical conditions.
Snow Agent is not intended as a one-off, but as a way to build out deeper agentic functions across airport operations. Next steps include expanding Snow Agent to additional airports, and extending the solution to new use cases such as: Foreign Object Debris (FOD) detection, wildlife monitoring, and environmental and infrastructure monitoring.
Organizations like Aeropuertos Argentina are redefining how we think of ERP. Personally, I find these industry AI co-innovation projects much more compelling than the typical “copilot” productivity scenarios.
While we don’t have the full results from this first winter season yet, applying AI to mission critical projects, infused with real-time information, just hits me harder than generalized office productivity stories. A story worth tracking indeed…