As a child, Sebastián Flores wanted to be Tony Stark. A fan of Marvel and Star Wars for as long as he can remember, he chose to study engineering with a clear goal: to learn enough to one day assemble the Iron Man suit.
He didn’t make it, he jokes, or at least not yet: it could be a long-term hobby. What it did build was a platform that, with artificial intelligence, seeks to protect workers on construction sites.
He defines himself as a restless and anxious person, always looking for something to do. Before starting a business, he says, he was much more of a dreamer: he spent his time designing application mockups.
Today you know what it costs to execute an idea. That distance between dreaming and executing runs through its history, and also the way it thinks about technology: without shortcuts and with its feet on the job.
From security cameras to construction
Neos was not born for risk prevention. At Prototypes, an event at the Adolfo Ibáñez University, Sebastián Flores, Benjamín Salas and Santiago Sfeir, fellow undergraduates and graduate students, with eight or nine years of work together, presented an image recognition solution applied to security cameras.
The presentation was heard by an entrepreneur (Maria Paz Gillet), who suggested that they bring this technology to construction and connected them with a construction company, who told them their greatest fears. The main one: accidents.
They accepted the challenge, and from there came the current platform, which is no longer just cameras: it has a document manager, a conditions manager and an accreditation module with solutions, each with small layers of artificial intelligence.
- What did you see in risk prevention that made you think there was an opportunity?
“The construction industry is not very digitalized: many things are carried out on paper, and digitizing does not mean that it is done well. There is a lot of valuable data that is not used…
…There are companies that offer you to digitize everything, but you arrive and they are free texts, generic forms that do not adapt to the reality of the company”
“The opportunity is to offer something that really moves the customer’s needle: that adapts to their company, that listens to them and that also provides innovation. That customer remains loyal, because that is what they were looking for and couldn’t find it”
To understand this problem, he insists, you have to be on the ground: its end user is not the construction company or the managers, but the preventionists, the field managers, the foremen and the workers.
At the beginning, he admits, they didn’t know anything about risk prevention. Today they have a platform that clients can customize to suit them.
First the bases, then the artificial intelligence
The Neos recipe has a precise order: first digitize well, manage change, organize and standardize data; only then, apply artificial intelligence. “Not the other way around,” he summarizes.
Sebastión Flores would have loved to reach a market already prepared for AI, but this is not the case. And even when AI works, he maintains, it cannot be left alone in an area where people’s safety is at stake.
- How do you see your responsibility in building an AI that helps make decisions in such a sensitive area?
“We have always believed that you will never be able to automate everything with AI: there has to be a layer of manual verification. If we are talking about a critical document, AI can help you 80%, but there always has to be human validation behind it. If it saves you work, great, but it always has to be reviewed by people.”
Regarding the priority that companies give to security, he responds that it depends on the culture: some only want to fulfill a role; Others have it in their roots, and those are the ones he prefers to work with.
His argument for the former is economic: a fatal accident can stop an entire project, and that is why, he says, investing in prevention is much more profitable than assuming the costs later.
- What would have to happen in the region for these technologies to become a standard?
“First we have to have good data governance. And that my competition also has it, please, because when we receive the transition from them to us it is a disaster. Then we have to optimize, do the digital transformation and only then apply artificial intelligence, so that the tool works like a car on a highway, not like one on a dirt road full of potholes”
Persist, persist, persist
Sebastián does not hide that the path had dark moments. There was a point where he and his partners were on the verge of giving up, not knowing if they were truly adding value to the industry or just to a company. What kept them going, he says, is something they share with humor.
- What has been the most difficult moment you have experienced as an entrepreneur?
“Not knowing what will happen tomorrow. There was a moment when we wanted to throw in the towel, because we didn’t know if we were really giving value to the industry. But we have this issue of persisting. We are nerds, we really like anime and Marvel, and one of the values that anime brings is to persist, persist, persist until we achieve it.”
He recognizes that it’s nearly impossible to tune out: Sometimes you’re in the middle of a set at the gym and you’re already pitching an idea for a client to a partner. His only rule is that he doesn’t work on Saturdays: he sleeps late, goes out on his bike, and resumes work on Sunday at five in the afternoon.
For next year, the goal is to add more clients and consolidate the Neos name: anonymity, he admits, cost them, and the competition has already learned about the technology they offer and began to develop something similar.
- What advice would you give to that Sebastián who was leaving his master’s degree with the idea of creating Neos?
“Keep going. Have a focus and pursue it. The path has many obstacles, but with persistence you will achieve it”
If he could change something, he says, it would be to have known the industry better before leaving, to arrive with more contacts. Regarding the Iron Man suit, however, he does not lose hope.
Meanwhile, build another kind of protection: one that, with patience, order and without rushing artificial intelligence, seeks to ensure that workers return home healthy.
