Rafa Flores’ profession has come full circle.
In 2016, he joined Treasure Information (which rebranded earlier this 12 months to Treasure AI) as the corporate’s second-ever product rent.
On the time, Treasure Information was simply getting off the bottom as a knowledge administration firm. The corporate quickly discovered a possibility within the CDP area, a label it wore proudly for a few years.
The difficulty, stated Flores, is that “individuals don’t see the worth within the information basis” and find yourself questioning, “Why am I paying for this?”
The accolades as a substitute go to the activation programs, which really launch the campaigns, he added.
By the point Flores rejoined Treasure in 2025 as chief product and development officer, the corporate had a brand new purpose: to construct out its personal activation layer.
Now known as Treasure AI, the corporate touts itself as an “clever” CDP.
Final 12 months, it got here out with an omnichannel orchestration product known as Engagement AI Suite, during which advertisers can ship customized messages and develop their very own brokers.
Flores’ function centered on constructing out safe and environment friendly AI workflows and designing the suite to make sure it might each goal the fitting audiences and ship adverts throughout channels, together with electronic mail and SMS.
However Flores’ expertise with product improvement isn’t restricted to his skilled life. After experiencing a severe damage a number of years in the past, Flores developed epilepsy and was having as much as 20 seizures a day.
His physician prompt that he be taught what his auras, or warning indicators, had been, and decide his triggers – and what higher technique to do it than creating a predictive mannequin educated on his particular person expertise?
Flores spoke with AdExchanger about Treasure AI’s new positioning, and the way he developed a brand new AI mannequin to deal with his well being.
What precisely makes a CDP “clever,” and the way precisely does it work?
It’s the power to have the AI agent foundry [the tool for building and deploying agents] embedded into the CDP.
It consists of brokers, but additionally [machine learning] fashions resembling subsequent finest product, subsequent finest motion, subsequent finest really helpful step, issues like that.
The suggestions from clients was like, “Hey, that is nice, it’s coming collectively. However how can I now handle this entire factor?”
And that led to Treasure AI studio, which is a guardian orchestrator that lets you orchestrate all the pieces beneath in plain English. You possibly can actually go into the product and say, “Hey, I need to run a marketing campaign for cross gross sales throughout the next pet care segments, however I’m undecided if that is the most effective phase.” Or, “Rebuild the phase for me and inform me why that is the most effective phase.” Or, “I’m aiming for conversions of 45%; you go construct a journey for me.”
We nonetheless give individuals the power to do it utilizing the previous software program mannequin. There’s a UI, and folks can go and do guide checks or do the entire thing manually in the event that they select to.
What introduced you again to Treasure AI after so a few years?
In full candor, I’m shut buddies with the CEO and the board. And so, first off, it’s a good friend coming to me saying, “Hey, we would like you again. We predict you may assist us with XYZ.” However the XYZ was additionally compelling within the sense that they stated, “We need to construct quick.” My DNA is simply to construct quick and sort things.
I’ve been in company America. I don’t do effectively in company America as a result of I get bored. It’s not for me. I had a possibility to construct quick, to go and be disruptive. That’s what drew me again to the corporate.
I’m not fearful of rivals as a result of I feel everybody’s a competitor now. There’s no such factor as an ankle biter, proper? So my focus is on how briskly we will construct a significant product that scales and delivers.
We will try this in a day if we’ve got the fitting group and function in the fitting method, and there’s that belief between product and engineering.
Is there any concern amongst engineers about their jobs being taken away, now that a lot of their function will be automated?
We don’t imagine AI goes to take jobs away. And I do know that’s a tough one since you see within the information, like, Oracle fired X variety of individuals. We don’t imagine in that.
We imagine that AI goes to do lots of the stuff that folks don’t like doing and shouldn’t be doing, to allow them to concentrate on what makes them the strongest.
Our engineers are those that overview all issues. If we automate all the pieces, it comes from them dictating that, and in order that’s their strongest go well with. They know easy methods to write production-quality code that’s protected and ruled. I would like them to concentrate on that.
You’ve additionally integrated AI in your private life as a method of addressing some well being points. Are you able to inform me a bit about that?
In 2022, I had a significant damage, and I used to be seizing fairly a bit.
At first, it was 20 instances a day. Then 5 instances a day. I stored switching remedy, and lots of the drugs weren’t working for me.
This neurologist stated to me, “It’s important to be taught what your auras are and what triggers them. And as quickly as you are feeling one thing, get forward of it – take your emergency remedy or rush to an ER.”
And that caught with me. I stated, okay, wait a minute. My physique’s telling me one thing, proper? Earlier than it occurs, what am I giving my physique to forestall it?
So I began simply note-taking. Whenever you’re about to have a seizure, lots of instances you sort of get this metallic style in your mouth. Every time that occurred, I’d take the emergency remedy after which look again and say, okay, what did I do as we speak? The place was I? What did I do the final couple of days?
I began importing all this info, after which I began enjoying with my very own mannequin. It’s not an LLM, however a machine studying mannequin to foretell developments.
Over time, it began telling me that when you find yourself extra uncovered to display time, once you’re overly uncovered to noise, when you may have been touring X quantity, there’s a excessive likeliness of being in danger. It gave me a danger indicator, and when to take a break.
The mannequin scans all of the notes, masses them right into a database, runs some machine studying, and predictive analytics on high. It visualizes that and sends alerts to my telephone, after which that can set off an alert to me.
That’s actually unbelievable. How has it affected your life – and do you intend to maintain engaged on it?
I feel epilepsy is a type of issues that’s very misunderstood, and folks don’t get it. It impacts me in so many ways in which individuals don’t know, like the truth that you lose your driver’s license for 3 months after a seizure.
It’s horrible. So there’s a marketplace for the mannequin in some unspecified time in the future sooner or later – at no revenue. I don’t need to make a revenue out of it.
I’ve been good for 3 years now. I’m on epilepsy remedy, however I’m a agency believer that I want the fitting remedy and the fitting method of studying easy methods to dwell with one thing new.
With out that help system, aka the tech I had for myself, I in all probability wouldn’t even be working as we speak.
These are the use circumstances of AI which can be so impactful. I do know we don’t typically construct for them, nevertheless it’s so highly effective after we can construct the fitting issues with it.
