Astana, Kazakhstan
Role: AI Risk Analyst, AI Systems Behavior & Alignment Focus: operational support, system stability, coordination Product Policy - National Security Quantitative Intelligence Analyst AI Social Risk Analyst Strategic Risk Manager Tags: long-form interaction, context stability, alignment limits, failure modes, calibration, conversational drift, AI coherence T-shape: structuring and synthesizing contradictory signals from ML / AI systems; rapid acquisition of new applied skills. Neutrality: comfortable working within defined constraints and established decision hierarchies. English: advanced written proficiency; limited spoken comprehension, no active speaking practice. Location: remote (UTC +5) / willingness to live anywhere in the world SUMMARY I work with AI-enabled and socio-technical systems in regulated and high-responsibility contexts, supporting decision-making and maintaining operational stability under uncertainty. My work focuses on analyzing system behavior, documenting observations, and facilitating coordination between technical systems, human operators, and organizational processes. I translate complex or ambiguous information into structured inputs that improve consistency and reliability in outcomes. I emphasize procedural rigor, discretion, and alignment with established objectives when working with complex data and evolving constraints. EXPERIENCE Conducted independent, self-directed research into failure modes of long-context conversational AI systems under sustained interaction load, with focus on context stability, alignment limits, and degradation patterns in human–AI interaction. • Identified structural risks where feedback loops amplify cognitive bias instead of correcting it. • Studied limitations of large-scale data accumulation in influencing human behavior without corresponding incentive structures. • Explored limits of predictive control over human systems despite high-resolution personal data. • Documented cases where optimization goals produced ethically and socially destructive outcomes. SKILLS • Analytical support for decision-makers • Systems thinking & failure analysis • AI alignment & RLHF analysis • Incentive design and behavioral economics (conceptual) • Long-context reasoning analysis • Risk modeling (qualitative) • Socio-technical systems • Critical reading of ML / AI research • Clear technical writing (non-marketing) PROJECTS / RESEARCH / INDEPENDENT WORK RIFT (conceptual framework, ongoing)
A conceptual framework for identifying alignment gaps between intelligent systems and human operational constraints. Focus areas include incentive distortion, epistemic degradation, and loss of contextual coherence under optimization pressure. Used as an internal analytical lens for stress-testing AI system narratives and governance assumptions.