Neurologyca, a company focused on building human-centric artificial intelligence, has announced the launch of Kopernica, an AI platform designed to interpret a wide spectrum of human emotions.
Kopernica leverages multimodal inputs such as live audio and video, combined with behavioral intelligence, enabling AI applications to "comprehend" human emotions. The platform identifies stress and anxiety by analyzing subtle changes in tone of voice, facial expressions, behavioral cues, and mental states.
Through these markers, the company aims to enhance emotional intelligence, allowing AI models to better understand humans across various environments.
"Current AI systems can understand what we say but not how we feel," said Juan Granja, co-founder and CEO.
To achieve this, Kopernica employs a unique 3D pattern recognition modeling capability that monitors over 790 reference points on the human body—seven times more than existing market solutions. For audio, the underlying model captures not just words but also speech patterns like intonation, rhythm, and prosody.
Neurologyca stated that unlike platforms heavily reliant on manually labeled public datasets, Kopernica is trained on decades of neuroscience research with a strong scientific foundation.
Underpinning its signal processing component, the model uses a deep learning framework consisting of 10 processing layers capable of evaluating up to 90 emotion classifications. This enables it to determine complex states such as motivation, stress, attention, and cognitive load.
"With Kopernica, we’ve created a human context layer that allows these systems to not only capture nuanced human emotions but also respond with empathy, adapt their behavior, and truly enhance human-machine relationships," Granja explained.
The company describes Kopernica as functioning as an infrastructure layer, referred to as an "Emotion Operating System," designed to work alongside large language models or other AI agents and applications.
All processing occurs locally on devices, using anonymized insights to avoid storing or sharing any identifiable user data without consent, according to Neurologyca. This approach emphasizes privacy-by-design principles in developing its infrastructure.
The potential misuse of emotion-detecting AI has long been a concern for privacy and legal experts.
"If mismanaged or improperly supervised, emotion AI could cause significant harm to individuals and expose companies to substantial legal risks," warned Lena Kemp, lead attorney at LK Law, writing in Business Law Today. "It collects and processes highly sensitive personal data tied to intimate emotions, with the potential to manipulate and influence consumer decision-making processes."
Existing global regulations, such as the EU's AI Act, already prohibit the use of AI-driven emotion detection and prediction in policing, workplaces, and classrooms, with only a few exceptions allowed for medical treatment and safety purposes.
According to Neurologyca, when used responsibly, Kopernica could revolutionize how people interact with AI, such as OpenAI’s ChatGPT. Emotional enhancements would enable AI agents and large language models to adapt to human emotions, adjusting tone and rhythm based on user input; health applications could monitor emotional and cognitive states; and media content platforms could detect mood and recommend improved entertainment options.
Using the same infrastructure, Kopernica could also be applied in clinical systems to flag early indicators of cognitive stress or stroke risk.
Neurologyca, a Europe-based company, recently opened its first office in San Francisco. The company reports strong interest from U.S.-based AI leaders, health innovators, and infrastructure partners. Kopernica is currently accessible to select partners, with broader availability planned for the second half of 2025.