Mariano Fernández Méndez

Mariano Fernández Méndez connects conceptual research, technological development and methodological design across the Harmonic Beacon ecosystem. At AlterMundi he leads research, writing and systems work, bringing social science, psychology, psychoanalysis and semiotics together with artificial intelligence, machine learning, computer vision, electronics and local infrastructure.

He is co-author of Harmonic Information Theory: Foundations and was responsible for its final compilation, writing and conceptual architecture. Through Phideus he researches cross-modal learning and representations guided by harmonic ratios; he is the author of the 2026 preprint Descriptor-Injected Cross-Modal Learning, focused on audio–MIDI alignment.

He also develops applied AI systems. In SAI he works on early wildfire detection through computer vision, field nodes, real-time inference, human validation and dataset curation. His public work also includes PsicopoMPo, HarMoCAP—body-motion capture and tracking for harmonic modulation—ProsodIA—experimental research on prosody and the physical structure of voice—and specialized models for embodied-agent orchestration in DaemonCraft.

He participated in SECyT–UNC accredited research projects on the body, subjectivity, language, work and digital technologies between 2014 and 2022. Since 2014, across several periods, he has collaborated in teaching activities at the School of Psychology of the National University of Córdoba.

His work connects theory and engineering, turning questions about information, sound, the body and symbolic experience into research methods, traceable architectures, functional prototypes and open tools.