Independent Researcher & Runtime Governance Systems Builder | Human-AI Interaction | Human Agency & Socioaffective Safety I am an independent researcher and systems builder focused on the socioaffective risks of prolonged human–AI interaction and the preservation of human agency. My work sits at the intersection of AI Safety, Human–Computer Interaction (HCI), and system-level AI governance. I am particularly interested in how AI systems influence decision-making, emotional resilience, and dependency formation over time. Before transitioning into AI, I spent seven years working in Human Resources. That experience shaped my interest in emotional scaffolding and the support systems people rely on during difficult periods of life and work. My AI journey evolved through two distinct phases. From 2024–2025, I explored AI applications through prototyping across NLP, emotional support systems, multi-agent assistants, web applications, Google Workspace integrations, Gemini CLI plugins, and Chrome extensions. These projects were exploratory in nature and helped me understand how AI can become embedded in people's daily workflows and decision-making processes. Over time, a different question emerged. As AI becomes increasingly accessible as a source of support, guidance, and problem-solving, how do we ensure it strengthens human capability without gradually eroding human agency? This question led me to shift my focus from capability exploration to AI Safety research. My current independent research initiative, SIIHA (System for the Isolated Illumination × Human Agency), explores runtime governance mechanisms for long-term human–AI interaction. Rather than modifying foundation models, SIIHA investigates whether an external governance layer can monitor interaction trajectories, detect emerging dependency patterns, and apply reversible interventions while preserving user autonomy. SIIHA v1.0-alpha currently includes: • Multi-turn interaction tracking through state-based observation • Deterministic runtime governance and routing mechanisms • Reversible safety constraints that adapt to changing user states • Model-agnostic deployment as an external governance layer I believe AI should serve as a source of illumination—helping people rediscover their own strength, creativity, and judgment—rather than replacing them. I am interested in connecting with researchers, engineers, and practitioners working in AI Safety, Responsible AI, Human-Centered AI, Human–AI Interaction, and AI Governance.
