AI Agents: The Good, the Bad, and the Ugly

Yup! AI agents are now a thing. They receive inputs like clients' claims, bills, or any unstructured information and do stuff with that. 24/7.

They can update your internal CRM, ERP, software or spreadsheet. Most people digest the main idea well.

So, where's the challenge, by chance?

• AI Readiness
• Implementation
• And the famous... "ok, who's gonna be driving this after we launch it"?

Yes, here I will post the summarized version; are you hungry for more, have a look at my other links:

• Longer Article in my blog (In English): https://weidemann.tech/the-role-of-ai-agents-in-customer-support/
• Longer Article in Forbes México (In Spanish): https://forbes.com.mx/impacto-de-la-inteligencia-artificial-en-la-atencion-al-cliente-en-mexico/

Now... back at it

AI READINESS

AI Readiness is a simple point. We all may hear, "I'm looking to have AI in our {insert here operation that it's today's pain}

But... can it be solved nowadays by a human? I remember a project that wanted to gather information from collaborators via WhatsApp and update SAP software with that info in real-time. That sounds lovely, right? So, I asked for a video of how a human gathers this information. They replied that they don't have any single WhatsApp number from their collaborators (after months of telling my team that they had updated, structured information about their collaborators....)

IMPLEMENTATION

Well, this is a good one. Did your team ever create a flow chart? If the answer is no... You first need to write a written version of your process. I've heard in many innovation projects the statement, "We know how to do it; it's a well-defined process."

I required someone in my team to create the flowchart of the process. And not being able to draw three nodes, as there were already more than 30 "it depends" answers to the questions about the process...

I always like to say that when I hear, "It depends on a well-defined variable," it needs a bifurcation (we know it in tech as IF). But when I hear "It depends on the situation," the answer is more related to a cognitive part of the business that you haven't learned deep enough about your problem yet.

And the famous... "ok, who's gonna be driving this after we launch it"?

Here's a fantastic thing—pure psychological. I have seen people who want to become the "cutting-edge tech company" in their sector as an OKR.

How do you obtain that OKR? "Let's hire interns." For some reason, the budget for well-qualified talent is so tight that for some leaders, if the title says "innovation / digital transformation," they deduce it's an "intern" task.

Results? No results. But I still wonder how these leaders surprise themselves. "How could this ever happen? We got the best interns."... (SARCASM DETECTOR!?? )

What's Next? Up-skilling and Re-skilling the talent whose job description is now performed by AI agents. I will go into this in my next article! Stay tuned!