Interview to Erik Cambria
Founder of SenticNet
As artificial intelligence rapidly reshapes entire industries, one of the greatest challenges for businesses today is turning powerful algorithms into practical, scalable tools. From no-code platforms to sentiment-aware APIs, the future of AI lies not just in innovation—but in accessibility.
To better understand how companies can bridge the gap between complex AI models and real-world solutions, we spoke with Erik Cambria, founder of SenticNet and one of the leading voices in affective computing. From training machines to read emotions to the future of explainable AI, here’s how he sees the evolution of AI in the B2B space.
As an expert in AI, how do you approach the challenge of turning complex AI models into practical, user-friendly applications or services?
Turning AI into something tangible—like an app or a business tool—is no small feat. Deep neural networks, the core of most AI systems, require massive amounts of data and processing power. They’re not easy to package.
That said, things are changing. Thanks to no-code platforms and AI-as-a-Service (AIaaS) solutions, businesses can now plug into powerful AI tools without building them from scratch. One such example is SenticNet’s own API offering, which provides sentiment analysis services for B2B companies through a ready-to-use platform. It’s a clear sign that AI is becoming more modular, accessible, and business-friendly.
How can artificial intelligence be applied to sentiment analysis, and what does it take to train AI systems to understand emotions in a more nuanced, qualitative way?
Most sentiment analysis tools rely on training data to label content as either “positive” or “negative.” While that’s a helpful baseline, it often overlooks the complexity and nuance of real human emotions. We use neurosymbolic AI, a hybrid approach that integrates generative models, such as LLMs, with structured semantic frameworks, such as knowledge graphs.
This allows our system to move beyond surface-level keywords and understand deeper layers of meaning—capturing tone, context, and subtle emotional cues, even in ambiguous or emotionally layered text.
At the center of this approach is the Hourglass of Emotions, a multidimensional model that not only detects how people feel but also uncovers the underlying emotional drivers—offering insights that are both rich and actionable.
“We use neurosymbolic Al, a hybrid approach that integrates generative models, such as LLMs, with structured semantic frameworks, such as knowledge graphs.”
Erik Cambria
Founder of SenticNet
How do you think AI is changing the B2B landscape and what do you see as the future challenges?
AI is no longer just a “nice to have”—it’s becoming a business essential. I see companies racing to integrate AI into their operations, and for many, the quickest way in is through plug-and-play AIaaS tools.
But beneath the surface, there’s a deeper shift unfolding. As adoption grows, so does the demand for AI talent—and I’ve noticed that many businesses are struggling to find or train the right people. That said, I’m optimistic. With the rise of accessible online courses and upskilling programs, I believe this talent gap will gradually close.
The bigger long-term challenge, in my view, is explainability. In today’s data-driven world, it’s not enough for AI to deliver accurate results—it also needs to explain how it got there. Especially in B2B, clients demand transparency. They want to understand the logic behind every classification so they can fine-tune their models and make smarter decisions. And I completely agree with that need.