How AI Is Complementing Human Work

In recent months, I have spent a lot of time studying and understanding where to apply artificial intelligence, using ServiceNow’s slogan: “Put AI to work for humans.” At first, it might sound like pure marketing, but it is actually the absolute truth: the companies and individuals who understand this first and put AI Agents to work for people will gain a major competitive advantage.

And this advantage will not only be economic, it will also be operational, strategic, and experiential for both employees and customers.

But first, we need to understand what these agents are, how we can use them, and what level of maturity companies need in order to use them effectively.

The AI Agent is the evolution of what we used to know as Generative AI.

Gen AI (like ChatGPT, for example) summarizes, assists, and works as a co-pilot: it can, for example, summarize a meeting and assign tasks with responsible owners and bullet points. It can also generate emails, documents, and proposals or respond to inquiries contextually, freeing up the time of human teams.

Then, based on those summaries, predefined business rules, and a knowledge base ( previously made decisions, for example), Gen AI can offer possible paths for a question or problem — like chatbots that direct you to department A or B, person A or B, depending on the case.

Now we arrive at AI Agents, the real stars of the moment. And this is where I believe humans can benefit greatly — gaining time and productivity.

The more specific the agent is, the more useful it becomes. We are not talking about general AI, but digital experts that “understand” a specific task.

Even though ROI is talked about a lot, the real return of AI is often seen in the form of time and productivity rather than direct monetary gains from a business case.

These agents understand a microprocess (think of each agent as a specialist in a specific task) or a subprocess within a broader process. For example, imagine a large pharmaceutical distributor in Mexico with many internal processes, but one in particular directly affects its revenue and its quality and level of service: onboarding pharmacies as new clients.

Before fully digitizing, it could take up to 60 days for a pharmacy to be ready to purchase medications. Even though this company’s mission is to deliver medication within 24 hours anywhere in Mexico, this process made it impossible.

Two years ago, they decided to implement the ServiceNow platform, digitizing the flow from form filling, document receipt and analysis, certification verification, expiration dates, and authorized medication types to credit analysis.

Once that analysis is approved, the pharmacy can become a client. But being a new client, what is the credit amount in pesos, and for which products?

This is where we decided to include an AI Agent: a “Super Credit Analyst” is being developed to analyze this final part of the process. This agent will cross-reference previously validated information (documents, certifications, among others) and, based on business rules and public data, suggest to the human in charge the most appropriate credit limit and payment terms.

A microprocess that used to take two days is now resolved in minutes.

Who wins with this? The human.

The human’s role is still to approve or reject the suggestion — the final decision is human — but much of the work is already done by the agent. That frees up time for higher-impact tasks and reduces human error in repetitive tasks.

That is “putting AI to work for humans.”

As I mentioned in my last article, I firmly believe in artificial intelligence. I am sure that we humans will be the main beneficiaries.

I do not believe AI is here to replace us, rather, when used correctly, it will save us time and make us much more productive.

I recently saw this study by IBM, and it left a very clear message: In Mexico, 45% of AI investments already show a positive ROI. 

Often, when talking about tech projects, we expect a clear financial ROI. But with AI, the return is often in time, productivity, and accuracy.

The big challenge I see today in Latin America, and particularly in Mexico, is that many companies still do not have digitalized processes.

Large investments have been made in record-keeping systems (like ERPs, HCMs, and so on), but these systems are not integrated. They need platforms like ServiceNow to connect them and build workflows between them. Only then will it be possible to automate with AI Agents, who can also interact with other AI assistants and move almost entire processes forward, leaving only final approval in human hands.

In short, it is not just about implementing AI but about preparing the ground with connected processes, clean data, and clear goals.

The future is going to be very interesting. Let us not fear AI — it is here to make life easier.

At ServiceNow, we are already challenging several of our long-term clients to identify a process or subprocess where we can implement an AI Agent. It is a collaborative effort, and there are already cases where clients have started to see success and returns.

Do you have a repetitive, slow, or critical process that still depends 100% on people? Let us talk. I am sure an AI Agent can help improve that task.