Cognition‑aware AI for Humans and Organizations

Here's what we discovered: AI fails not because it lacks intelligence, but because it doesn't understand how humans actually think and feel. We're building something different.

Most AI treats every user the same—like a perfectly rational decision-maker who just needs more information. But real people aren't like that. We catastrophize. We get stuck in all-or-nothing thinking. We make decisions based on values that seem irrational but make perfect sense to us.

So we're teaching AI to recognize these patterns. When someone says "this project will be a disaster," our system doesn't just offer generic reassurance. It recognizes catastrophizing, understands the underlying fear, and helps reframe the situation in a way that actually resonates—because it knows what this specific person values and how they prefer to process difficult situations.

This isn't just therapy-speak. We've built measurable systems that track real outcomes: how often teams reach better decisions, how much cognitive load decreases in conversations, how trust builds between people who communicate differently. Because if AI is going to live in our most important conversations—in boardrooms, classrooms, and moments of personal struggle—it needs to understand not just what we're saying, but how we're thinking.

Cognition-aware AI is not simply a set of tools; it is a research philosophy. Where much of AI research aims to make models faster, larger, or more accurate, our philosophy is to make them more humanly aligned. We treat cognition itself—distortions, traits, reasoning styles—not as noise to be ignored but as an infrastructure layer for the next generation of AI.

This philosophy shapes every application we pursue. In education and coaching, cognition-aware systems act not as answer engines but as companions in reasoning. By drawing on frameworks like Acceptance & Commitment Therapy (ACT), they help learners and professionals surface distorted thinking, reframe unhelpful patterns, and sustain motivation.

In HR and organizational design, the challenge is collective. Teams often lose clarity when communication styles clash, when bias creeps in unnoticed, or when hidden assumptions block progress. Cognition-aware copilots detect these patterns and provide timely signals: highlighting when a conversation is caught in all-or-nothing framing, or when contributions are being systematically undervalued.

In enterprise environments, cognition-aware AI determines whether copilots remain curiosities or become indispensable infrastructure. Standard copilots often fail because they feel generic, opaque, or psychologically tone-deaf. By contrast, cognition-aware copilots embed reasoning frameworks, adapt guidance to personal traits, and explain their interventions in human-legible terms.

Across these domains, the principle remains consistent: cognition-aware AI makes individuals think more clearly, teams work with fewer breakdowns, and enterprises adopt AI with confidence. What distinguishes this research is not only its applications, but the philosophy that drives them: AI must align with the structure of human cognition if it is to achieve both scientific rigor and social adoption.

Our work is peer-reviewed and published at leading venues in natural language processing and affective computing. Recent highlights include:

We are building a research program that is both scientifically rigorous and practically indispensable. In every case, collaboration means advancing a new paradigm where cognition is treated as infrastructure for AI.