Rafa Flores’ career has come full circle.
In 2016, he joined Treasure Data (rebranded to Treasure AI earlier this year) as the company’s second product hire.
At that time, Treasure Data was just getting off the ground as a data management company. The company quickly saw an opportunity in the CDP space, a label it had proudly carried for many years.
The problem, Flores said, is that “people don’t understand the value of data infrastructure” and end up asking, “Why should I pay for this?”
Instead, praise goes to the activation system that actually starts the campaign, he added.
By the time Flores returned to Treasure as chief product and growth officer in 2025, the company had a new goal: building its own activation layer.
The company, now called Treasure AI, bills itself as an “intelligent” CDP.
Last year, it launched an omnichannel orchestration product called Engagement AI Suite that allows advertisers to deliver personalized messages and develop their own agents.
Flores’ role focused on building secure and efficient AI workflows and designing the suite to target the right audience and deliver ads across channels such as email and SMS.
However, Mr. Flores’ product development experience is not limited to his professional life. After suffering a serious injury several years ago, Flores developed epilepsy and began having up to 20 seizures a day.
His doctor suggested learning about his aura, what his warning signs are, and identifying his triggers. What better way to do that than by developing a predictive model trained on his personal experience?
Flores spoke to AdExchanger about Treasure AI’s new positioning and how it has developed new AI models to address health.
What exactly makes a CDP “intelligent” and how exactly does it work?
The ability to have a foundry of AI agents [the tool for building and deploying agents] Embedded in CDP.
This includes agents, but [machine learning] Models such as next best product, next best action, next best recommended step, etc.
Customer feedback was as follows: “This is great. Things are going well. But how can we manage this whole thing?”
And that led to Treasure AI Studio. This is the parent orchestrator that allows you to orchestrate everything in plain English. You can literally think about a product and say, “I’d like to run a cross-sale campaign across the following pet care segments, but I’m not sure if this is the best segment.” Or, “Rebuild the segment and tell me why this is the best segment for you.” Or, “I’m aiming for 45% conversion. You build the journey for me.”
We’re still giving people the ability to do that using older software models. There is a UI that allows you to manually check and run everything manually if you wish.
What brought you back to Treasure AI after all these years?
Frankly, I’m close friends with the CEO and the board. So, first of all, a friend came to me and said, “Hey, we’d like you to come back. We think you can help us with XYZ.” However, XYZ was also persuasive in the sense that we wanted to make it quickly. My DNA is to build quickly and fix things.
I have worked for an American company. I don’t do well in corporate America because I get bored. It’s not for me. There was an opportunity to build quickly and take disruptive action. That’s what drew me back to the company.
I’m not afraid of competitors because I think everyone is a competitor now. There’s no such thing as someone biting your ankle, right? So my focus is on how quickly we can build meaningful products that are scalable and deliverable.
If you have the right team, operate the right way, and have trust between product and engineering, you can do it in a day.
Are there concerns among engineers that their jobs will be taken away, now that many of their roles can be automated?
We don’t believe AI will take away jobs. I know this is a tough question because I’ve seen news like Oracle laying off X number of people. We don’t believe that.
We believe that AI will do many of the things humans don’t want to do or shouldn’t do, so humans can focus on what makes them the best.
Our engineers review everything. When we automate everything, it’s at their command and that becomes their strongest suit. They know how to write secure, controlled, production-quality code. I want them to focus on that.
You’re also incorporating AI into your personal life as a way to deal with health issues. Can you tell me a little bit about that?
In 2022, I had a major injury and was having quite a bit of cramp.
At first it was about 20 times a day. Then 5 times a day. I kept changing meds, but many meds didn’t work for me.
This neurologist told me, “You have to learn what your aura is and what triggers it. And as soon as you feel something, be proactive and take emergency medicine or rush to the ER.”
And that stuck with me. I said, “Okay, give me a second.” My body is telling me something, right? What can you give your body to prevent it before it happens?
So I just started taking notes. People often feel a metallic taste in their mouth when a seizure is about to occur. Every time that happens, I take my emergency medicine and then turn around and say, “What did I do today?” where was i? What have I been up to these past few days?
I started uploading all this information and then started playing with my model. This is not an LLM, but a machine learning model for predicting trends.
Over time, we’ve learned that if you spend too much time looking at a screen, being overexposed to noise, or traveling X amount of distance, you’re more likely to be at risk. It gave me a risk indicator and told me when to take a break.
The model scans all notes, loads them into a database, and performs machine learning and predictive analysis. Visualizing it sends a signal to my phone, which triggers an alert to me.
That’s really incredible. How has it impacted your life and do you plan to continue working on it?
I think epilepsy is one of those diseases that is very misunderstood and that people don’t understand. It affects me in many ways that people don’t know, including losing my driver’s license for three months after the attack.
That’s terrible. Therefore, at some point in the future, a market for this model will exist without profit. I don’t want to profit from it.
I’ve been doing well for three years now. I am on medication for epilepsy and I strongly believe that I need the right medication and The right way to learn how to live with new things.
If it wasn’t for that support system, the technology that I had for myself, I probably wouldn’t still be working.
These are highly impactful AI use cases. I know we don’t often build for them, but when we can build something right with it, it’s very powerful.
