In an era when climate-driven disasters are accelerating faster than response systems can adapt, SkyTL stands at the frontier of predictive technology and emergency intelligence. Founded by Rocio Frej Vitalle, an aerospace engineer turned entrepreneur, the company is redefining how we forecast and respond to wildfires and other large-scale environmental threats.
Drawing from her years in government contracting and her frustration with how slowly innovation reached the field, Rocio built SkyTL to close a critical real-time gap between research-grade climate modeling and actionable decision-making. Today, her team’s work is helping emergency responders anticipate risk, protect communities, and operate with unprecedented speed and clarity.
In this conversation with Eqvista, Rocio shares the story behind SkyTL’s creation, the science driving its flagship product WindTL, and her vision for how technology can transform disaster resilience worldwide
Rocio, you started your career in aerospace engineering and have worked with government contractors. What made you decide to leave that world and start SkyTL?
I came from the world of government contracting, where progress was slow and technology took many years to see light. I wanted to move faster—and build something that could, too. SkyTL was born from that frustration and a belief that disaster response shouldn’t lag behind the speed of risk. We had science. We had the tech. What we lacked was a way to operationalize it in the field. So I left federal contracting and built SkyTL to close that gap between advanced modeling and real-time decisions.
Wildfires are getting worse and more unpredictable because of climate change. How do you build models that work when fire behavior is changing in ways we haven’t seen before?
We don’t rely on a single model. There are many new technologies and Earth observation systems that capture data and have mitigation capabilities we didn’t have before. Our system fuses not only atmospheric data, fuel modeling, infrastructure maps, and satellite imagery but also the actions of risk reduction or mitigations and processes it through both physics-based models and machine learning. By integrating real-time data—from drone feeds to field sensors—we update forecasts and provide the best course of action options to reduce risk before the fire and mitigate and active fire every few minutes. This hybrid approach helps us adapt to new fire behaviors as they evolve, not after the fact.
Your flagship product, WindTL, has already demonstrated remarkable real-world impact. The Thompson Fire case in 2024 is particularly striking; your system predicted embers would cross a river within two hours while other forecasting tools said the fire was contained. How did your system get that right?
Forecasting spot fires due to embers has been a huge challenge in the past, and a critical one, as embers are responsible for 90% of structure losses during wind drive fires. The good news is that this new decision making approach we bring to the table can help us solve questions that we could not before, such as ember transport. To understand ember behavior, we need highly localized weather, vegetation, and mitigation efforts information.
Our new approach allows us to bring all these components, so we can forecast where embers are going and if they will ignite new fires. During the Thompson Fire, other tools saw containment; we saw a wind corridor capable of carrying embers across a natural barrier. That forecast gave responders critical lead time to adjust their plans and protect a community before it burned.
We’re curious about your hybrid processing engine that combines fire modeling tools from academia, research labs, and your proprietary models. How do you handle model disagreement? When these models disagree with each other, how do you decide which one to trust?
Disagreement is expected—and valuable. Sometimes, the truth lies not in choosing one model, but in understanding why they diverge and where each model performs best. That’s where our proprietary technology helps arbitrate and deliver actionable insights.
Your partnership with Trident Sensing for real-time aerial infrared imaging is fascinating. This seems to be moving the wildfire response from reactive to truly anticipatory. What does “real-time” actually mean in this context? Are we talking seconds, minutes?
It depends on the scenario. For the analysis of aerial imagery, aircraft observations during a fire are rarely input into forecast modeling. And if so, it happens after the aircraft lands and it is a very manual process. In this case, real-time means in a matter of minutes. We’re integrating sensor data from collection to actionable insight through a highly automated process. That includes airborne IR imaging, UAS feeds, and field sensors. It’s not just fast—it’s fast enough to inform resource staging, route planning, and even evacuation orders while events are still unfolding.
You’re providing advanced predictive tools to firefighters and emergency managers who work under extreme pressure. How do you make your system simple enough to use without losing the important details?
Our system is built on one principle: emergency responders don’t need to be highly trained fire analysts to run and understand forecasts, as it happens today. We translate complex data into simple, color-coded maps and clear alerts. Think: green, yellow, red zones. Dashboards are intuitive, with embedded recommendations. We also integrate with tools they already use, like ArcGIS and TAK, so SkyTL becomes a natural extension of their workflow.
Congratulations on your recent $3.5 million funding round. What challenges do you anticipate in scaling from pre-seed to the next financing stages, and how do you plan to use these funds to overcome them?
Thank you! Over the next year, we are laser focused on scaling our software infrastructure, our partnerships, and gaining trust from users and partners. We’re investing in cloud infrastructure to handle larger datasets, expanding our partner network with cities and utilities, and ensuring every new customer and partner sees results fast. That builds trust. The funding also helps us scale our team without compromising our velocity or mission.
You’ve received support from NASA, NOAA, and Google for Startups, but according to recent data, SkyTL hasn’t raised traditional venture capital yet. That’s an interesting strategic choice in a space where many startups are raising large rounds. What’s your thinking/Philosophy on funding the company’s growth?
We’ve been deliberate about that. We are talking about national infrastructure resilience, so to create a significant increment in the state of the are, we had to invest in years of R&D, customer engagement, pilot programs, and input from experts in leading agencies such as NASA and NOAA. We knew that if we wanted to build something truly ground-braking about how we manage disasters, we needed to build a very robust platform that overcame the science and technical challenges others have come short of. We wanted our tech to prove itself in the field first. Now that we’ve done that, we’re open to strategic capital—but it has to align with our mission.
As a woman founder in aerospace and emergency management, both fields with very few women, what’s that experience been like for you, especially when it comes to fundraising and visibility?
That is very true. These are industries where you still walk into rooms and you’re the only woman. But I’ve also found allies and built a company that proves competence is louder than skepticism. Every successful deployment, every new partner—that’s visibility earned. And I’m committed to pulling others up as we grow.
Finally, looking three to five years out, where do you see SkyTL? Are you building toward a specific exit, or is this a long-term platform play? What does success look like for you personally and for the company?
SkyTL is a platform play. We’re building the operating system for environmental threat response. In 5 years, I want SkyTL to define the way we mitigate risk and respond to disasters. I want us to be the backbone for every team preparing for a disaster or responding to one, either as the modeling backbone of their current tools, or as their standalone tool, not just in the U.S., but globally. Personally, success means delivering life-saving intelligence before a disaster, not after. That’s the future we’re building—one deployment, one partnership, one forecast at a time.
