Paula Petcu, Co-Founder and CEO of Interhuman AI, joins Bright Founders Talk to share her journey at the intersection of technology and human behavior. With a background in computer science and a deep fascination with the human mind, Paula has built a career focused on understanding how people think and act.

In this conversation, she reflects on her path from consulting and corporate leadership to founding a company dedicated to decoding human behavior through AI. Interhuman AI aims to tackle one of the most complex challenges in tech today: modeling humans when we still struggle to fully understand ourselves.

Paula discusses how advances in AI, especially the rise of large language models, have opened new possibilities in this space. She also highlights the limitations of traditional data collection methods and the need for more nuanced, human-centered approaches. Drawing from her experience in digital health, she explains how real-world challenges inspired the creation of her company. This interview offers valuable insights into the future of AI, human understanding, and the bold ideas driving innovation forward.

Teaching AI to Understand Humans: Paula’s Bold Bet on Behavior

Paula’s career doesn’t follow a straight line—it weaves through consulting, corporate leadership, and deep tech, always circling one big question: how do humans actually work? From her early days as a computer scientist to leading roles in healthcare and startups, she kept coming back to the same fascination—the human brain. That curiosity eventually turned into something bigger: Interhuman AI, a company built around decoding behavior, emotions, and the subtle signals we barely notice ourselves.

But here’s the twist—Paula isn’t just building another AI tool. She’s tackling something far messier. “We don’t have a ground truth about ourselves as humans.” That’s the core challenge. Imagine trying to teach machines to understand people when people don’t fully understand themselves. It’s ambitious, slightly chaotic, and exactly the kind of problem that attracts a team of curious minds—behavioral scientists, AI experts, and researchers all trying to map the uncharted territory of human behavior.

The idea didn’t come out of nowhere. It was shaped by real-world frustration. While working in digital health, Paula saw firsthand how limited existing tools were—wearables, surveys, clinical data—all missing the most important layer: how people actually feel and behave in real life. You can’t capture body language or emotional nuance in a questionnaire. That realization pushed her to experiment, and when breakthroughs like ChatGPT hit the scene, everything clicked. Suddenly, AI wasn’t just analyzing text—it had the potential to see, hear, and understand. And for Paula, that opened the door to something much bigger.

Beyond Words: Teaching AI to Read Between the Lines

Words are just the surface—and Paula is determined to go deeper. Today’s AI models may be incredibly fluent, but they’re still missing a huge part of what makes us human. The tone of voice, the subtle shift in posture, the glance away when something feels off—these are the signals that carry real meaning. Paula and her team at Interhuman AI are building technology that doesn’t just listen to what we say, but actually observes how we say it, spotting the gaps between words and emotions.

That’s where things get really interesting. “Sometimes you say one thing, but your body says something else.” That contradiction—so natural to humans—is exactly what AI struggles with today. By combining video, audio, and text into one model, Paula’s team is trying to give machines a kind of social intelligence. Think of it like a digital therapist or coach that doesn’t just hear you, but reads the room. It’s not about replacing human intuition—it’s about augmenting it, especially in areas like healthcare, training, and communication.

But Paula is careful not to get carried away by the tech alone. There’s a constant balancing act between data and people, innovation and responsibility. The team tests their own product on themselves, exploring its limits, biases, and real-world impact before putting it out there. And the goal isn’t to replace humans—it’s to build AI that can actually collaborate with us. In Paula’s vision, the future isn’t humans vs. machines, but humans working alongside AI that finally understands more than just words.

From Signals to Sense: Paula’s Vision for Socially Intelligent AI

Building AI that understands people isn’t just a technical challenge—it’s also an educational one. Paula knows that before companies can fully adopt this kind of technology, they need to understand why it matters in the first place. Social intelligence in AI isn’t a “nice to have”—it’s the missing layer that determines whether these systems actually work in real human environments. And right now, a big part of her mission is simply helping the market catch up to that idea.

At the center of this push is their latest breakthrough: a model called Inter 1. “It captures video, audio, and text at the same time.” Not separately, not stitched together afterward—but processed in real time as a single stream of human behavior. That shift—from multimodal to truly omni—opens up a whole new level of accuracy and speed. It doesn’t just label what’s happening; it explains why. Raised eyebrows, a hesitant tone, a slight head tilt—suddenly, AI can point to the exact signals behind its conclusions, making it far more transparent and usable in real-world applications.

And the possibilities? They’re wide open—but Paula is staying focused. Instead of trying to conquer every industry at once, her team is zeroing in on areas where human interaction really matters, like sales training and coaching. Because at the end of the day, closing a deal isn’t just about numbers—it’s about reading the room, building trust, and responding in the moment. That’s exactly where socially intelligent AI could make the biggest difference.

Why AI Won’t Replace Human Connection—But Might Make It Stronger

Deals aren’t closed with spreadsheets—they’re closed with trust. Paula makes it clear: no matter how advanced technology becomes, human relationships will stay at the heart of business. Whether it’s referrals, personal networks, or just that gut feeling about someone, people still buy from people. And that’s exactly where her technology steps in—not to replace that connection, but to strengthen it. By helping sales teams understand both what they say and how they say it, AI can turn every conversation into a learning opportunity.

“It’s the relationship that you build.” That’s the real driver behind every successful sale—and Paula’s team is leaning into it. Imagine a new sales rep practicing pitches and getting instant feedback not just on their words, but on their confidence, tone, and body language. Maybe they said all the right things, but hesitated at the wrong moment—that’s the kind of insight traditional tools miss. With AI stepping in to highlight those subtle signals, onboarding becomes faster, feedback becomes sharper, and performance improves in ways that feel almost intuitive.

But Paula isn’t just thinking about business impact—she’s thinking about responsibility. There’s a clear line she won’t cross. No surveillance, no lie detection, no misuse of deeply personal data. Instead, the focus is on trust, transparency, and giving users control. In her view, AI should feel like a helpful partner, not an invisible observer. And for founders walking a similar path, her advice is refreshingly simple: stay passionate, trust yourself, and be ready for the doubts—because building something new was never supposed to be easy.