Sofia De la Mora Tostado is the founder of Intelital, an AI company building predictive intelligence systems powered by knowledge graphs. She is an AI engineer specializing in graph theory and network analysis, known for her work using knowledge graphs to dismantle human trafficking networks, which earned her the Sotero Prieto Prize, Mexico's highest mathematics award. She holds a bachelor's degree in applied mathematics, a bachelor's degree in political science, and a master's degree in data science from Columbia University.

How did your early work analyzing human trafficking networks shape how you understand complex systems today?

That work taught me to look for the structure behind complex situations, not just the surface signals. In human trafficking cases, understanding hidden relationships and behavioral patterns meant the difference between dismantling a network or missing it entirely. Graph theory allowed me to map connections that weren't obvious at first glance, revealing truths that surface data couldn't show. That experience fundamentally shaped how I approach intelligence work today. High-stakes decisions, whether about stopping trafficking or closing an enterprise deal, don't happen randomly. They emerge from people, timing, context, and relationships moving together in very specific ways. Learning to see those patterns early in my career gave me a framework I now apply to business intelligence at Intelital.

How has your identity as a young female engineer and immigrant shaped your journey from research to founding a company?

Being a young female engineer and immigrant in the United States means occupying a space where very few others exist. I speak openly about this because it's shaped everything about how I build. There are very few of us, and that means I've had to work harder to be taken seriously, especially when moving from research into entrepreneurship. Building credibility as an immigrant founder adds another layer of complexity.

But I don't view my identity as purely a challenge. It's given me a different lens. I see problems others miss because my experience is different. Being underestimated forces you to be sharper. You can't rely on assumptions or networks that naturally open doors. You have to prove value immediately. That discipline has shaped how I build Intelital, focused on delivering measurable impact from day one. I've learned that the fastest way to overcome bias is to be undeniably good at what you do.

How do your degrees in applied mathematics, political science, and data science come together in your work?

For me, these disciplines aren't separate. They converge in how I think about intelligence systems. Applied mathematics provides the rigor to model complex patterns and build predictive systems that actually work. Political science teaches me to understand power structures, incentives, and how decisions actually get made in organizations. Data science gives me the tools to extract meaning from noise and turn raw information into actionable intelligence.

Real intelligence isn't just prediction. It's understanding the context, relationships, and timing that shape why people act. That requires thinking across disciplines. When I build predictive models at Intelital, I'm not just optimizing algorithms. I'm modeling how humans make decisions under uncertainty, how organizational dynamics influence buying behavior, and how relationships shape outcomes. You can't do that well with just one lens.

What inspired you to move from engineering to founding Intelital?

After my academic work, I built a career as a data and AI engineer, but I kept noticing the same gap everywhere I looked. Enterprises were drowning in data but starving for real intelligence. Companies had dashboards full of signals, but no one could tell them who actually needed their product, why, or how to reach them. The turning point came when I spoke with over a hundred go-to-market leaders and heard the same frustration repeatedly. Their tools generated noise, not insight. They were guessing on their highest-stakes decisions.

That's when I decided to build Intelital. My technical background in graph theory and network analysis positioned me to solve what others treated as an automation problem but what I recognized as a prediction problem. Automation without accurate prediction just makes failure happen faster. What enterprises need is an intelligence layer that understands how their world actually works. My policy background helps me understand that these aren't just technical challenges. They're about understanding how organizations make decisions, what incentives drive behavior, and how to design systems that actually serve the people using them.

What does Intelital do and why does it matter?

Intelital builds predictive intelligence systems that are hyper personalized to each client. My models power high-stakes decisions across sales, investing, procurement, and recruiting. I chose to begin with enterprise sales because it is the function where unpredictability carries the highest cost. Conversion rates in cold outreach have collapsed as AI tools made it easy to send thousands of superficial messages, overwhelming prospects and burying real intent. To break through today, companies need perfect timing, meaningful personalization, and warm introductions. That is exactly what I deliver. My platform identifies who to talk to, why they need your product now, and the warmest path to reach them, turning go-to-market from a high-volume guessing game into a precise, high-conversion intelligence motion.

How are you disrupting the GTM industry?

Intelital is disrupting the industry by challenging the assumption that go-to-market is a workflow automation problem. I believe it is fundamentally a prediction problem. Automation without accurate prediction simply makes failure happen faster. Our approach is different because every client receives their own predictive model powered by a knowledge graph that maps every interaction across sales, marketing, and customer success. This lets us adapt the intelligence layer to the unique patterns of each organization rather than relying on one-size-fits-all scoring.

Our AI agents track complex, highly specific intent signals for each client, including behavioral patterns such as engagement around events, public conversations, changes in personnel, customer reactions on social platforms, and even relevant legal or market activity. Each signal has a weight that adjusts continuously with every win and loss, which means the predictive model improves over time and learns how each company actually buys. This focus on adaptive, personalized prediction enables me to deliver a level of precision in account and lead prioritization that traditional go-to-market tools cannot match.

How do you know the market needs what you're building?

I spoke with more than a hundred go-to-market leaders to understand how they make decisions and where their tools fall short. The pattern was always the same. Their platforms were generating dozens of shallow signals and long lists of leads they had to review manually, leaving them guessing about who actually needed their product. It became clear the market was missing true intelligence, not more notifications. When I showed teams a system that could identify who was ready to buy, why, and the warmest path to reach them, the response was immediate. They moved from curiosity to urgency and started paying right away. That level of pull was the strongest confirmation that I was solving a real, high-stakes problem.

How do you differentiate in a crowded GTM space?

Intelital differentiates by focusing on something most go-to-market platforms overlook: the network behind every decision. Sales does not happen because a score changes but because relationships, timing, and context align in very specific ways. We train a dedicated predictive model for each customer, allowing us to map the hidden connections and signals that shape real buying behavior and to reveal not only who is ready to buy but why and what path will actually work. Instead of giving teams more dashboards to interpret, our intelligence delivers clear next steps they can act on immediately. Leaders want trust, not noise, and by giving enterprises predictions that hold up in the real world, we are able to win high-value clients who want an intelligence partner rather than another tool to manage.

What is your long-term vision for Intelital?

To create the intelligence engine that supports every major decision inside an enterprise. We are starting with sales because it is the fastest way to prove the value of accurate prediction, but the underlying technology is designed to extend far beyond go-to-market. The same reasoning capabilities that identify buying readiness can also guide decisions about hiring, investing, partnerships, procurement, customer expansion, and risk. My goal is to build a system that understands how an organization moves, learns with it, and gives leaders clarity exactly when they need it. In the future, I want Intelital to become the resource executives rely on when the stakes are high and getting it right truly matters.

How does your graph theory background shape Intelital's technology?

Graph theory taught me to look for the structure behind complex situations, not just the surface signals. At Intelital, I apply the same way of thinking to business decisions. Buying readiness does not appear out of nowhere. It comes from people, timing, context, and relationships moving together. Using knowledge graphs helps me map those dynamics so I can understand where opportunity is forming and how teams can reach it. It is simply a practical way to make better predictions by understanding the real structure behind decisions. Instead of treating every lead as an isolated data point, we map the entire ecosystem of relationships, behaviors, and context that influences whether someone will actually buy. That architectural choice is what enables the level of precision we deliver.

Why did you join Techstars?

I joined Techstars because I wanted structure, speed, and a community that would challenge my thinking. I had a strong technical foundation but needed a clear, repeatable go-to-market motion, and Techstars gave me the discipline to find it. The program pushed me to run real discovery, test assumptions quickly, and stay focused on the problems that matter most to customers. The NYC cohort also offered access to founders and mentors who understand enterprise sales and AI, which made it the right environment for the stage I was in.

What lessons have shaped you most as a founder?

One of the biggest lessons has been the importance of charging from the beginning. Payment is the clearest way to know whether you are solving a real problem, and free products often hide the truth. I also learned to stay obsessed with the problem, not the solution. The more conversations I had with customers, the more I realized that real traction comes from curiosity, simplicity, and continuous experimentation. Every insight that has moved me forward came from listening closely, testing quickly, and staying focused on what genuinely creates value.

What advice would you give to other technical founders, especially women and immigrants, who are thinking about starting a company?

Start before you feel ready. The credentials, the perfect network, the polished pitch, none of that matters as much as solving a real problem for real people who will pay you for it. For women and immigrants especially, your outsider perspective is an advantage, not a limitation. You see gaps others miss because you've had to navigate systems that weren't built for you. Don't wait for permission or validation from gatekeepers. Build something undeniable, charge for it from day one, and let the results speak louder than any credential ever could. The path won't be easy, and you'll face bias that others don't. But the fastest way through that is to be so good at what you do that the work becomes impossible to ignore.

Read More