Introduction: AI Value Grows With Team Adoption

How many of you work in a company? Yeah.

So it's pretty fun to use AI. But if you are the only one using it, you are not using it properly.

Because if you work in a 10 -people company, not a huge company, and every person of the company is saving 10 hours with AI, it's great. You are going the right path.

But if you are the only one using it, You are the geek that is saving 15 hours. It's nothing compared 15 hours to 100 hours.

So what we do basically is, OK, we are going to work with the company to save time. This is not updated, but whatever. We have worked with many companies in Spain, helping them adopt AI, and saving a lot of time.

A Practical Framework for AI Adoption Maturity

Why Time Savings Get Harder to Measure Over Time

that part of six hours is only in the first month as a data and as a curiosity when you start working more and more time with AI it's get harder to identify the numbers of hours that you are saving for example I've been working with AI since the GPT release I wouldn't be able of knowing how much time I'm saving by using AI I don't know if that happened to you and if you know in what states of of AI use you are right now.

So I have built this framework so all of us can identify in a part of use AI.

Level 1–3: From Asking Questions to Copy‑Pasting Outputs

First of all, we have I ask AI about topics. Instead of searching on Google, I do that. Later, OK, I ask them to write me a blog post or email. and I copy and I paste it.

Later I don't have to explain everything always because it knows me and I tell generate a description of my company and it gave me something something perfect for example.

Level 4: AI With Context From Your Tools and Files

Later we have something that we are approaching to what Juan has shown that is it can read my systems my files my tools and answer through that.

How are we going with this project? Okay, I will enter your project management tool and answer according to that.

Level 5–6: Delegation, Automation, and Building AI-Native Workflows

But this is boring. This is not good for me. The magic of AI starts here for me.

That is, okay, I want you to do that. I want you to create one presentation according to all information.

I want you to develop this tool according to that. I want you to generate or update my website according to all the knowledge that our company has generated in the last month.

Meetings, messages, whatever. And the last one is, as Juan has shown, you build your own tools on that. Maybe they are not tools

with an interface. They could be tools that work with natural language. And if

you have heard that the revolution right now is that we are going to go to no interface era. Have you heard that? No, not many of you.

The “No-Interface” Direction: Natural-Language Control of Apps

So basically we are going to talk about Gmail, for example, that many of you know that the tool. What What do you need to enter your email?

If you can just have your AI tell you, you have 20 new emails. Only five need your answers. I will tell you, ask you questions, and according to that, I will answer.

For example, when I use my phone, I just tell him, OK, Siri, set a timer of 10 minutes. I don't go to, as an example of this happening right now, I don't go to the interface of clock and activate the timer.

So this While the time of AI is getting cheaper and is more adopt is going to happen more and more Okay, yeah really quickly.

Benchmarking the Room: Where Most People Actually Are

I mean we're going to vote to see the level of the room How many of you think are in the level one? One person two

Don't be shy level two level three 3, OK, 3, 4, level 4, 3, level 5, 3, 4, 5, level 6.

OK, we have here more people.

A Common Gap: Building Tools Without Letting AI Do the Work

Something that I have seen according to this framework is that some people are here, they are building tools, but they are not here, for example.

They are building tools, but they are not about letting the AI do the work. I just let you do that here so you can think about that.

First of all, when you use AI as, okay, do that and give me that, it's great. But the power, and I want to give you an example, is that it does all the work.

How Teams Ship Deliverables Without Opening Traditional Apps

In our team, team, what we have seen is that almost all the time, the people are working in Cloth or Codex or in video calls in Teams. They are not using any other thing, almost, because

if they want to create a PowerPoint, the PowerPoint of one I guess that it was built in Codex, This one, it was built in Clot. We didn't enter into PowerPoint to build the presentation. We just, from Clot, give it the instruction

on how it should work to do the presentation. Okay.

These are the levels that we have seen. It has the context, read your tools, and this is the good level that is okay.

How This Looks in Practice: A Consulting Project Workflow

For example, when we work with a consultant project, project, so you can see how we are working. We do interviews.

We record them and the way you're processing that interviews is just plot. Plot plus our human input. That is,

we have observed that these people or this person wasn't really happy when we said about this tool. Maybe it doesn't like.

So we give our information that we receive as humans that AI is not there yet.

Assessing Adoption in Your Company

Using a QR-Based Survey to Find Strengths and Weak Points

Okay, what I want you to do is scan this QR really quickly because we have take a guess on where you are. But right now we are going to see actually where you are in AI adoption.

This is a tool that we are basically working for. Okay, let's have the real number of adoption of the companies. If you translate it with Google Translator on your browser, it does.

I'm going to give a few minutes so you can, because what I want you to see is that after that, you are going to receive a number from 0 to 100, but that's not the interesting part. The interesting part is when you go down and you select to see per part of the adoption.

Beyond Usage: Compliance, Protocols, and Data Readiness

So it's not only that you are using a lot of AI, you are using it with compliance, you are using it with protocols, you are using it with all your data, is that the thing that it gets your attention?

So, I'm going to show you, and now I'm also going to show you how we have built that with Lobapol. I want to show you, really quickly, oh you are in process, okay, so the example, it could be, okay, so what you are going to see is something similar to that.

This is what, ah you are not seeing it. Why not? Okay.

When you finish, I want you to make your focus and read this part. Maybe you are using a lot of AI with really huge technology, but your internal resources are not working properly.

This is from a company that they were quite good in AI, but But they weren't working with resources internally.

So any of you have finished the QR questions? That want to raise their hands? Anyone has finished? OK, I'll give a few minutes more.

Tool Spotlight: Building Internal Apps Without Heavy Dev Ops

What “Lovable” Does and When It Helps

Who knows which tool is that? Someone take a guess. No, that's not Shell GPT. This is lovable.

For any of you that didn't understand when Juan said what is a pull request, that I assume that is many of you, lovable, it could be your tool. Because what it basically does, it helps you build whatever tool without entering in GitHub or any deploying part. part.

It centralizes all the deployment. So, we could come here, just prompt, I want whatever, and we will have the tool. So, for example, I want a tool for truck expenses in my company.

I think it's not going to work because I have run out of credits, but you could just say that and you could have a functional tool. So, I just wanted to show you that, okay, Okay, what you have seen is completely done in lovable.

What You Can Configure (Files, Database, Backend) Without Living in the UI

In lovable you have different parts. First of all, you can see all the parts of the files, you can see the part of the database, and you can see all the part internal,

but you don't need to come here. This part you don't need anything. You just need to use this part, and say, I want you to do this. Yeah, this could be the backend in a really visual way of seeing it.

Trade-Offs: Convenience vs. Terminal-Level Flexibility

So something interesting that we have been seeing before is that Juan said, okay, I want you to fix some vulnerabilities. Okay, you can come here, just say scan in dev or quick scan and try to fix all. And it's something similar.

But the handicap is that it's not so versatile than if you use the terminal and cloud code directly.

Conclusion: Use the Results to Decide Your Next AI Moves

So really quickly, because I think the value is in the questions, you can do many things with AI that I'm sure you are not doing.

If you don't want to share the results, don't worry, but you have there where is your weak point on AI usage. it and let me know