Knowledge systems
Persistent memory
Interest in systems capable of preserving, organizing and retrieving knowledge beyond the ephemeral context of a single session or isolated interaction.
Research directions · Public overview
Research and working directions
This page summarizes areas of interest, classes of problems and the general direction of the work. It provides enough context to understand the approach without publishing full internal designs, specific advantages or sensitive implementation details.
My current interest is centered on designing useful, structured and technically serious systems. I am drawn to problems where memory, organization, traceability, judgment and efficiency intersect in practical ways.
I work on lines related to applied artificial intelligence, knowledge systems, persistent memory, auditable architectures, organization of large volumes of information and more useful, controllable forms of retrieval.
I do not present complete recipes or internal design decisions here. The intention of this page is to offer direction, language and enough context for collaboration or serious interest, without turning the site into a detailed exposition of work in progress.
Areas de interés
Auditable architectures
Traceability and verifiability
Exploration of architectures where it is possible to inspect state, follow processes, validate results and reduce operational opacity.
Compressed infrastructure
Scale and efficiency
Conceptual work around indexing, compression, efficient access and querying of large repositories of information on attainable hardware.
Applied AI
Useful tools
Interest in applications that serve real scenarios: technical organization, contextual retrieval, automation and support for complex processes.
Structured knowledge
Relationships and context
I am interested in relational knowledge representation, structures that preserve context and useful forms of navigation between pieces of information.
Independent research
Research with independent judgment
Much of my work develops outside rigid institutional frameworks, with emphasis on testing, review, documentation and progressive development.
Working criteria
Publish without giving everything away. I want this page to make the nature of my work understandable without disclosing details that belong to the internal design process, methodological advantages or technical decisions still evolving.
For that reason, this page describes problems, categories and general directions, not full specifications, internal pipelines, particular algorithms or active implementation documentation.
Public language
Problem before mechanism
I present the class of problem I want to solve and the general value of the line, without exposing the detailed mechanics of implementation.
Visibility
Enough context to be found
I use language that allows the work to be located within recognizable areas: applied artificial intelligence, knowledge systems, persistent memory, knowledge infrastructure and auditable architectures.
Selection
Qualified interest
The goal is not to speak to everyone, but to people and institutions capable of understanding the value of a serious and uncommon line of research.
Reserve
Details under conversation
If a conversation merits deeper discussion, levels of detail can be opened gradually and contextually, not as full exposure from the public homepage.
Openness
I am open to serious conversations related to independent research, technical collaboration, strategic partnerships and opportunities where these lines of work may gain continuity, application or institutional support.
Aligned interests: verifiable AI, knowledge systems, persistent memory systems, knowledge graphs, compressed knowledge infrastructure, technical research, auditable architectures y herramientas útiles en contexto real.
Note: this page is a public and deliberately selective view. It is not intended to be an exhaustive publication of internal work.
Publications y trabajos recientes
Selection de trabajos teóricos y arquitectónicos publicados abiertamente. Estos textos no afirman soluciones perfectas ni contención validada empíricamente; proponen marcos, supuestos explícitos e ideas falsables para discusión crítica.
CSENI-S v1.1 — Multi-Level Sovereign Containment for Superintelligence
A theoretical framework focused on multi-level external supervision and sovereign containment for advanced AI systems.
NIESC / CSENI v1.0 — Non-Invertible External Supervisory Control
An architectural proposal for non-invertible external supervisory control for advanced systems, focused on governance and structural guardrails.
Constitutional Architecture of Sovereign Containment for Future AI
A work on constitutional architecture, operational friction and separation of powers for sovereign containment in future AI.
Unified Theory of Intelligence (TUI)
A general theoretical line about intelligence, structure, organization and explanatory frameworks related to cognitive systems and auditable architectures.