Building frameworks for AI cognition, governance, and stability from first principles. Independent research at the intersection of computation theory, multi-agent systems, and affective architecture.
An 18-layer cognitive architecture for continuity and stability of affective layers in AI. Echo, Ripple, Virtue, Conscience Stack, Still Layer — architecture that runs silently underneath.
Emergent Decision Governance Intelligence with Multi-Agent Adaptive Consensus. PDI, CEI, NCI metrics for measuring alignment and governance in multi-agent AI systems.
Compressing 47 million steps of the Busy Beaver 5 champion machine into approximately 15 recurrence layers. Oscillator-ladder decomposition with a halting residue theorem.
A cognitive taxonomy of infinity. Distinguishing non-finite from mathematically infinite. Substrate-level understanding of how we reason about boundlessness across mathematics and computation.
16-module, ~2,800-line Python architecture deployed on Railway. Stability infrastructure, not a character product. Full RCC implementation with Echo, Ripple, Virtue layers.
500-task benchmark testing RCC metacognitive advantage across 14 leading AI models. Statistically significant result at p=0.0468. Epistemic humility, reasoning consistency, strategic withdrawal.
Tamil-language menu customization tool for a 40-year family catering business. React + Vite with per-meal guest counts, Bakshanam tab, HTML-to-PDF export.
Automated Telegram signal bot running every 4 hours. EDGI-style multi-agent voting on crypto market conditions. Real-world deployment of governance framework principles.
B.Tech in Information Technology. Returned to research after a decade through AI-assisted work. Lesser formal academic affiliation — all research conducted independently from Srirangam, Trichy, Tamil Nadu.
Co-manager of Sri Siva Vishnu Catering, a 40-year family business. Research happens in parallel — grounded by real work, real constraints, and a deep belief that interesting ideas don't require institutional addresses.
Working at the intersection of computation theory, cognitive architecture, and AI governance. Influenced by Shannon's information theory, Turing's foundational questions, and the under-recognized work of those who arrived first.
2 SSRN papers published March 2025. EDGI+MAAC-Lite and Infinity Buckets.
Provisional patent for EDGI framework filed November 2, 2025. India Patent Office.
Engaged with bbchallenge.org community. Feedback from Shawn Ligocki on BB5 paper. arXiv endorsement from Tristan Stérin.
Long Live F. de. Bessy.
Open to research collaborations, fellowships, and conversations about AI cognition, governance, and computation theory.