Meet the Team

Bezawit designs AI systems that enhance human-like interactions. Her research includes generating realistic, context-sensitive smiles to improve rapport and communication in client interactions with AI systems. She develops NLP and deep-learning algorithms (Python, TensorFlow, Keras) to enhance AI agents’ backchannel smiles, enabling broader statistical analysis, and optimizes performance by analyzing smile intensity and duration.

Kevin’s work sits at the intersection of cognition, learning behaviors, signal processing, and AI systems— aiming to better mimic the human brain than traditional neural networks. He advances the Forward-Forward algorithm with a single forward pass enhanced by signal processing to power a network of lightweight Forward Neurons, an artificial frontal cortex. Using Python and PyTorch, his team explores signal methods, algorithmic optimizations, and longer-term goals around reasoning, bias, learning disorders, and early childhood learning.