About ALICE
ALICE -- Adaptive Learning Intelligence via Cognitive Engines -- is a web-based knowledge platform that hosts structured bodies of work as graphs of linked units and attaches configurable AI agents to each unit. It treats knowledge as a landscape to be explored rather than a sequence to be completed; every reader's path through it is genuinely their own; and the software takes part in the exchange instead of only storing and delivering content.
ALICE is developed and operated by Attruvera Technologies, Inc., a Delaware corporation headquartered in San Antonio, Texas, and a research spinout of The University of Texas at San Antonio. The platform has an open core and proprietary components. Each customer institution is served as its own tenant, with its own database, file storage and signing keys, and remains the controller of its data.
That premise was developed over three decades of theoretical and applied work at the intersection of mathematics, engineering, literary theory, and the study of hypertext as a communicative medium -- long before the industry had a framework for it.
Configurations
One platform, several content profiles. The same software runs each; what differs is the content and the roles.
Education. Instructors author courses as linked lessons with learning objectives and rubrics. Students read, submit work, converse with a course-scoped tutor, and receive AI-assisted formative feedback that instructors review before release. Instructors grade and manage sections, rosters and messaging. ALICE serves The University of Texas at San Antonio in courses offered under university research programs.
Vela. An invention-disclosure and portfolio-management service for technology-transfer offices. Inventors submit disclosures through a guided form; the office reviews, tracks and manages the resulting portfolio. Vela runs at vela.attruvera.ai.
Theoretical Foundations
ALICE is built on three intellectual pillars that are structural, not decorative.
Roland Barthes' concept of the lexia -- the atomic unit of a text whose meaning is relational rather than self-contained -- defines the basic unit of knowledge in ALICE. A lesson in ALICE is not a module to be completed; it is a lexia whose significance shifts depending on which other lexias the learner has encountered and in what order.
Gilles Deleuze and Félix Guattari's rhizome model defines the architecture of the knowledge graph. There is no root, no trunk, no prescribed hierarchy. Every concept is a potential entry point. Every connection is traversable in both directions. The path a learner follows is genuinely their own because the platform was designed to have no single correct path.
The third pillar is a theoretical claim articulated by Prof. Juan B. Gutiérrez in 2000, before the technology existed to fully realize it: that the medium of communication does not merely carry the message -- it acts on it. In the context of adaptive learning with AI agents, today it is a precise technical description of what AI agents actually do. The agents in ALICE do not retrieve or summarize content. They engage with it as active participants in the communicative event, producing meaning in the encounter rather than transmitting it from a repository.
A Thirty-Year Genealogy
The intellectual lineage of ALICE begins with Literatronica, a literary hypertext engine developed in the mid-1990s by Prof. Juan B. Gutiérrez and funded by cultural institutions in Colombia and Spain. Literatronica was not an educational platform -- it was a working literary system that demonstrated the viability of rhizomatic navigation as an authoring and reading architecture.
That work attracted serious scholarly attention. Nine doctoral dissertations, three master's theses, and five peer-reviewed articles by researchers in Spain, Germany, the United States, Iceland, Colombia, and Argentina examined Literatronica's theoretical and technical contributions. That body of secondary literature existed before ALICE had its first grant.
When the first NSF award arrived, the architecture was extended into education. The platform became ALICE: Adaptive Learning for Interdisciplinary Collaborative Environments. The lexia and the rhizome remained, now serving the construction and navigation of course knowledge graphs.
The current generation adds a third architectural layer that the original theory anticipated but could not yet build: cognitive agents that are themselves active communicative media. The platform is now ALICE: Adaptive Learning Intelligence via Cognitive Engines.
Architecture
ALICE is the knowledge platform: the content graph, the lexia model, the rhizomatic navigation engine and the multilingual content layer. Three further layers, each independently valuable, complete the Attruvera stack.
Max is the multi-agent harmonization layer. When several AI agents work on a task, Max coordinates their outputs -- resolving conflicts, enforcing consistency, and producing one harmonized result with a traceable decision record.
Hudo is the attribution layer. It binds each AI output to its contributing agents, sources and reasoning chain, so that the question every regulated setting must be able to answer -- who, or what, produced this, and on what basis -- always has an answer.
TVA, the Trusted Verification Authority, is the cryptographic foundation. It creates a tamper-evident record of AI interactions -- what was asked, what was produced, what was verified -- that anyone can check independently.
Funding
The development of ALICE has been sustained across three decades by funding agencies on three continents. Peer review at this scale and over this duration is not a credential. It is a record.
USA: National Science Foundation (NSF). Award #2518973, 2026-2029.
USA: National Institute of General Medical Sciences (NIH NIGMS). Award #1R25GM151182, 2023-2028.
USA: National Science Foundation (NSF). Award #1645325, 2016-2019.
Spain: Ministerio de Industria y Turismo de España. Sub-award from contract TSI-070300-2008-67, 2008-2009.
Colombia: Instituto Distrital de Cultura y Turismo de Bogotá. Award #IDCT-410/1998.
Colombia: Instituto Distrital de Cultura y Turismo de Bogotá. Award #IDCT-514/1997.
Colombia: Colcultura -- Instituto Colombiano de Cultura, now the Ministerio Colombiano de Cultura. Award #COLCULTURA-SECAB 014/1996.