Dewey: Humanitarians AI's Educational AI Framework

Naming an AI framework after a philosopher is a statement of intent. Humanitarians AI named its educational AI orchestration platform Dewey, after John Dewey, the progressive educational philosopher whose experiential learning approach shaped how generations of educators think about teaching. The framework's design, a turn-of-the-century professor with an inquisitive gaze and a thoughtful demeanor, is meant to embody that same unwavering commitment to educational innovation.

What the Dewey framework actually does

Dewey is an open-source platform for educational intelligence, and its job is orchestration: coordinating a set of specialized AI agents so they work together rather than as disconnected tools. Just as John Dewey revolutionized education with his experiential learning philosophy, integrating multiple educational approaches into a coherent whole, the framework's orchestration layer is designed to make sure its various agents collaborate effectively to create personalized, interactive learning experiences rather than operating as separate, siloed products.

Why a character, not just a platform name

The choice to give the framework a face, a designed character rather than an abstract product name, is deliberate. The Dewey character is presented as the embodiment of the framework itself: a scholarly, turn-of-the-century professor whose aesthetic is meant to suggest careful, inquisitive teaching rather than an anonymous piece of software. That design choice mirrors the framework's larger goal, bringing AI-driven Socratic teaching methods into computational education in a way that still feels grounded in a long tradition of thoughtful pedagogy rather than a purely technical product.

The agents Dewey currently coordinates

The framework already supports a range of groundbreaking projects across education. The Medhavi projects build intelligent AI books that adapt to individual learners, addressing what the framework describes as a "chicken and egg" problem in education by guiding conversations through dynamic Socratic dialogue rather than static content delivery. Ada is a calculus learning assistant that provides step-by-step guidance through complex mathematical concepts. Newton is a physics chatbot that incorporates dynamic simulations to help visualize scientific principles. Grace is an algorithms chatbot that simulates computational processes with interactive visualizations. Botspeak focuses on teaching effective communication with AI through practical tutorials, treating AI fluency itself as a skill worth deliberately teaching. Case Crackers enhances management education with customized case studies and simulations, extending the same orchestration approach into business education rather than confining it to STEM subjects.

Each of these agents targets a different subject and a different kind of learner, from someone struggling through an introductory calculus problem to a management student working through a case study. What they share is the underlying design philosophy: guide the learner through the reasoning rather than simply hand over an answer.

Why orchestration matters more than any single agent

The individual agents are each useful on their own, but the framework's real contribution is the coordination layer that sits above them. Rather than treating each tool as a standalone product, Dewey positions them as specialized instruments within one educational intelligence system, echoing the way John Dewey's own philosophy wove together multiple pedagogical approaches instead of relying on a single method. That orchestration is what allows the framework to move toward personalized, interactive experiences instead of just delivering the same content to every learner, and it is what lets the framework discover, in practice, which approaches to educational agent computing actually work.

Key takeaways

• Dewey is Humanitarians AI's open-source framework for orchestrating specialized educational AI agents.
• It is named after philosopher John Dewey and designed around his experiential learning philosophy.
• Current agents under the framework include Medhavi (adaptive AI books), Ada (calculus), Newton (physics), Grace (algorithms), Botspeak (AI communication), and Case Crackers (management case studies).
• The framework's central value is its orchestration layer, which coordinates agents to work together rather than as isolated tools.
• Medhavi's AI books use dynamic Socratic dialogue to adapt to individual learners.

Who this is for

Dewey is relevant to educators, developers, and volunteers interested in building or contributing to agentic educational AI systems through Humanitarians AI. By joining the community, contributors help develop these agentic systems, which are already enhancing personalized learning and working toward democratizing access to quality education. Anyone curious about how personalized, AI-driven learning tools can be coordinated into a single coherent platform can learn more and get involved at humanitarians.ai/dewey.