San Diego, California.- The nearly 22,000 attendees of the Cisco Live conference appropriate the center of San Diego, on the banks of the bay, for a few days. They literally take hotels, streets, convention center and even Petco Park, the parents stadium, to make a party with a DJ and even with the rock group, The Killers.
And that is nothing for Cisco, a company of 269,000 million dollars (MDD) and that in 2024 obtained profits for 53,800 million pesos. The information technologies giant, founded in 1984, changed the way companies operate and communicate with each other. From routers, data centers, wireless networks and even software, cloud, cybersecurity and cameras for videoconferences, the huge consortium has shaped telecommunications since the end of the 20th century and the first quarter of this.
Among the ads made in the Cisco Live were several software and hardware products, but the one that most caught the attention has not even come to the market and no one else (except for some Cisco customers) has tried it. It is a solution made with agricultural artificial intelligence, also known as agentic ai.
“Last year we talked about chatbots, and now we talked about how to manage this new workforce: humans and agents,” said Chuck Robbins, president and CEO of Cisco. The product in question is Cisco Canvas AI, a kind of artificial intelligence service (AI) with steroids, much more powerful than the generative AI we use in Chatgpt or Gemini, capable of performing complex tasks, sectioning them in parts, obtaining information from other sources, all with high autonomy and almost without human intervention.
You may be interested: The Japanese company that seeks to conquer Latin America
«AI agents?
But what is AI AGENTIVA or AGENTIC AI? According to the definition of chatgpt: “It refers to AI systems that act as autonomous agents capable of making decisions, planning and executing actions to achieve objectives, often in complex and dynamic environments. The word agentic comes from agencythat is, the ability to act on its own. ”
“Agents are those parts of software that are handled by a life cycle, which will be being deployed in the processes and communicating between systems by definition, already with little or no human interaction and will be automating specific tasks,” explains José Alberto Márquez, head of technology and transformation for PWC Mexico.
For example, when there is a network problem or with a company platform, an employee lifts a ticket that is sent to the organization’s area, there the specialist analyzes it and, depending on the complexity of the problem and the workload it has, can solve it in hours or up to days.
With Canvas AI, companies can administer their network infrastructure proactively and preventively, which would save IT professionals for a long time, operating from a single platform and on a single screen.
Christopher Sánchez, professor at Egade Business School at Technological of Monterrey and A IA expert, reminds us of how we have reached this technology: “We start with the predictive AI, which is like the ancient AI; the most common AI is currently generative, but we are already moving towards the ia of agents. What is the difference? That they can do more things autonomously. Research, write a report, send me the key points and then write everything (the article) and make an image ‘”.
Now read: Millionaires in the New Era Trump
With this description, it sounds that AI AI can remove the work from this reporter and also replace many more people in various industries. But before becoming paranoid, let’s see what this can do for us.
In the specific case of Cisco and Canvas AI, this platform is intended to administer the networks and solve problems, something that can be very complex and routine for information technologies professionals.
Ramón Viñals, director of Engineering Solutions in Latin America of Cisco, explains that Canvas AI allows us to work from a single dashboard with several sources of information to administer network infrastructure and solve specific problems. “The solution is based on the cloud, it is an agent agent, and the more it has, this agent will communicate with the others (…) but you have to know how to do the prompt and ask the right question. ”
Economic potential
In 2024, the agentic AI had a market value of 5.2 billion dollars (MDD) in 2024, according to a market of Market.us, but in 10 years it will have a valuation of 196,600 MDD. That is, every year it will have an average growth rate of 43.8% and will increase its value up to 40 times.
“The agentic AI market is consolidating as an area of great interest within the technological sector, driven by its potential to improve efficiency and innovation in various industries. The main technology companies are invested strongly in the development of these systems, recognizing their ability to transform operations through the automation of complex processes and decision making. This is promoting a new wave of applications in fields that cover finances and health Supply chain and customer service management, ”says the report.
We recommend you: Mexico aspires to be the epicenter of innovation
In addition, Market.us points out that the main force that this technology is promoting is the need to improve productivity and operational efficiency. “Companies seek to reduce the human error and the necessary time for data processing and decision making.”
On the other hand, the Genai and Agentic ai do not work separately, the agents of Ia use generative, so the combination of both will bring a new era unsuspected productivity.
According to a recent McKinsey report, the generative AI technologies will add between 2.6 billion and 4.4 billion dollars of value in all commercial sectors and reduce the total amount of work required by all employees between 50 and 70%.

Agents and humans working
Given the legitimate concern that, the more the AI penetrates in the operations of the companies, it can replace more jobs, there are different visions. Christopher Sánchez, from Egade Business School, says that companies will first see if an AI can do a job, or if a human is needed, “or if we can do 50, 70, 80% of some job with AI.”
The academic highlights the need to take into account labor impacts. For example, more traditional industries have to think very well how their way of working with people will change in a new era of AI. Students who leave universities will not compete only with people in the labor market, but also with AI agents. “We will not be able to save all the works, that’s 100% sure,” said Sánchez.
We suggest you: Pemex owes more than 700 million parso from Carlos Slim Grupo Cars
On the other hand, there are those who see that technology does not have to be used expressly to move people. Carlos Pereira, customer experience in Cisco, recalls that much of the tasks that people do today are tedious, repetitive and that can be automated.
“We in Cisco are not thinking of putting AI agents to reduce the task, and with that reducing the number of people. It is not the plan. All we do is just help the human so that the tasks they have to do are more valued or that bring more value to the business,” he says.
A new (and unequal) was productivity
Organizations that integrate the AI successfully to their business will find many more ways to grow and take advantage of their resources, something that will be known as a new era of productivity.
Globally, industries with the greatest capacity to adopt AI are experiencing a three -time growth in the income generated by employee, according to the AI barometer in the 2025 work world, PWC, of PWC.
“The AI is quickly changing the skills required for collaborators. It is key to identify the skills gaps through data and define a Upskilling and Reskilling action plan to help reduce them,” says the report. In Mexico, companies are in an early AI adoption phase …
Christopher Sánchez warns that countries and organizations that dominate AI can have a great competitive advantage over others, even, may seem unfair. “We are going to see very important increases in global productivity, but we are going to see more, for example, in the United States, in China, then in Europe, Canada, Australia and then in the global south. And Mexico is part of that global south. But as in Mexico we do not have these tools, we will not have the same increases in productivity.”
Maybe you are interested: PayPal World: The platform that will connect more than 2 billion consumers
The academic also highlights the disadvantage of Mexico and Spanish -speaking countries in general in AI, because between 0.15 and 1% of the great language models (LLM) has content in Spanish, which makes it more difficult to enable agentic ai in these contexts. “And that for me is very dangerous, it is something that we have to focus very well because AI is to help people grow in an individual, as well as business. But if we are not focused on the most basic things, for example, the data that represents me, which represent you, we cannot compete within the country either,” warns.
Ethical and cybersecurity concerns
Another concern about AI AGENTIVA is about what so much autonomy should be given to the agents to carry out certain objectives, and that the objectives they make are aligned with those of humans.
There is also the possibility that cybercriminals can take advantage of vulnerabilities to attack objectives of companies or organizations. Matt Cauldield, vice president of identity product in Cisco, says that a “Zero Trust” philosophy must be adopted to enable ia agents in organizations.
“When we talk about zero trust (Zero trust) For agents, we refer to how to provide minimal privileges between agents and resources. That is what is braking companies to adopt agerentic intelligence: they want to give access to agents, but they don’t simply want to put them on the Internet, let them act without control and give access to email input trays without restrictions, ”he says.
José Márquez, of PWC, emphasizes that, with the incorporation of agriculture in business processes, one of the main challenges will be risk management and regulatory compliance (GRC: Governance, Risk, and Compliance).
Lee: Analysts expect dollar closing 2025 below 20 pesos
“With the adoption of agricultural artificial intelligence, companies must redesign their processes to guarantee traceability, audit and compliance, understanding not only what decisions are made, but why the AI takes them,” he says.
The agentic ai has the power to multiply productivity and transform industries, but also the risk of expanding the gaps between which they can play this new game … and those who are barely learning their rules.