プロフィール詳細

プロジェクトを作成

★★★★★

☆☆☆☆☆

★★★★★

☆☆☆☆☆

MinGi K.に依頼

South Korea

Computational Simulation Researcher | Digital Twin, AI Benchmarking, Failure Analysis & Candidate Prioritization

プロフィール概要

Kim Min-Gi – Independent Computational Researcher Computational Simulation | Digital Twins | AI Benchmarking | Failure Analysis | Candidate Prioritization PROFILE SUMMARY I am an independent computational researcher focused on simulation-driven hypothesis testing, complex-system modeling, digital twins, AI benchmarking, failure analysis, and candidate prioritization. I help researchers, startups, and technical teams transform complex ideas into testable computational models. My work focuses on a central question: Why do complex systems survive, collapse, adapt, and recover? I study whether common structural patterns may appear across biological systems, artificial intelligence, software systems, industrial infrastructure, and other complex networks. I am available for small and medium-sized computational research projects, exploratory simulations, proof-of-concept studies, benchmark design, and hypothesis stress-testing. SERVICES I can support projects involving: • Custom Python-based simulation • Digital twin prototype development • Complex-system behavior modeling • Failure and collapse analysis • Recovery-path analysis • Sensitivity and intervention testing • Candidate screening and prioritization • AI and RAG benchmarking • Comparative scenario analysis • Research hypothesis stress-testing • Data analysis and visualization • Technical research reporting Typical deliverables may include: • Python simulation code • Reproducible experiment scripts • CSV result files • Benchmark tables • Sensitivity analysis • Candidate rankings • Comparative scenario reports • Graphs and visualizations • Technical research summaries RESEARCH APPROACH My computational workflow generally follows: Problem Definition → Feature Encoding → System Modeling → Scenario Generation → Perturbation Testing → Failure Analysis → Recovery Analysis → Candidate Ranking → Validation Recommendations The objective is not simply to produce a simulation, but to identify: • Which variables matter most • Where instability begins • Which failures propagate • Which interventions improve recovery • Which candidate configurations deserve further testing • Which assumptions should be experimentally challenged SELECTED RESEARCH AREAS 1. OrganOS and Survival-Oriented Computational Architecture I am developing OrganOS, an experimental computational framework for reducing large candidate spaces into small active candidate sets while studying survival, collapse, recovery, and adaptive routing. Selected internal computational benchmark results include: • Experiments on candidate spaces up to 100 million items • Active candidate sets reduced to small survivor sets in selected configurations • Compression ratios exceeding 8,000,000× in specific large-scale experiments • Evaluation across multiple retrieval and benchmark settings Research topics include: • Thermal Memory • Echo-based propagation • Risk Memory • Survival-oriented selection • Collapse detection • Recovery routing • Adaptive governance systems These results are computational research findings and should not be interpreted as universal performance guarantees. 2. AI-Driven Molecular Candidate Prioritization I conduct exploratory computational research on reducing large molecular candidate spaces into smaller high-priority sets for further investigation. Selected internal benchmark experiments include: • ChEMBL: 100,000 → 17 active candidates • PCBA: 68,547 → 17 active candidates • HIV: 41,120 → 17 active candidates • QM9: 133,885 → 17 active candidates Research outputs may include: • Candidate prioritization reports • Candidate rejection analysis • Comparative ranking • Explainable screening logic • Follow-up validation recommendations Important: These are computational prioritization studies, not clinical validation or proof of therapeutic efficacy. 3. Biological and Disease-System Modeling I explore biological systems as interacting networks rather than isolated variables. Current research themes include: • Inflammation • Immune regulation • Leptin resistance • Insulin resistance • Adipocyte stress • Hypoxia signaling • Lactate accumulation • Acidity response • Cellular survival networks The objective is to generate testable hypotheses about system-level interactions, intervention points, and failure or recovery pathways. This work is exploratory computational research and does not replace laboratory, medical, or clinical validation. 4. Advanced Materials and Candidate Discovery I investigate computational approaches for generating, filtering, and prioritizing candidate material configurations. Research domains include: • Thermal materials • Radiation-related materials • Cryogenic systems • High-temperature structures • Aerospace-related material concepts • Multi-objective candidate screening Potential project outputs include: • Candidate comparison matrices • Multi-variable ranking • Sensitivity analysis • Failure-condition mapping • Experimental validation priorities 5. Security and Software Resilience I research collapse-aware approaches to software and infrastructure analysis. Topics include: • Risk-node detection • Failure propagation • Guard placement strategies • Recovery-path analysis • Large-scale code analysis • Software resilience modeling The objective is defensive analysis, robustness improvement, and failure prevention. 6. Industrial Digital Twins I develop exploratory digital-twin and simulation models for studying: • Risk accumulation • Failure propagation • Cost escalation • Recovery efficiency • Operational stability • Intervention scenarios • Long-term system behavior Potential applications include industrial systems, infrastructure, data centers, and complex operational networks. SELECTED PROJECT TYPES I am particularly interested in projects such as: • “Can you simulate whether this research idea is plausible?” • “Which variable is most likely to destabilize this system?” • “Where does failure begin and how does it propagate?” • “Which intervention should be tested first?” • “Can you compare multiple scenarios before physical experiments?” • “Can you reduce a large candidate set into a smaller priority list?” • “Can you build an exploratory digital twin?” • “Can you benchmark this AI or retrieval architecture?” • “Can you identify assumptions that may cause this model to fail?” RESEARCH VISION My long-term research goal is to identify common principles governing survival, collapse, adaptation, and recovery across complex systems. The central hypothesis is that systems may remain stable when critical flows of energy, information, resources, and control are maintained, while collapse may emerge when these flows become disrupted, overloaded, fragmented, or trapped in reinforcing failure loops. I investigate whether related structural patterns may appear across: • Biological systems • Artificial intelligence • Software systems • Industrial infrastructure • Ecological networks • Other complex adaptive systems COLLABORATION I am open to: • Small paid simulation projects • Exploratory computational studies • Research collaborations • Startup proof-of-concept projects • Digital twin prototypes • Benchmark development • Hypothesis stress-testing • Candidate prioritization studies For project discussions, please provide: 1. The research question 2. Available data 3. Variables or constraints 4. Expected output 5. Desired timeline I can then propose a computational research plan, simulation scope, deliverables, and validation strategy. IMPORTANT LIMITATIONS Computational models depend on assumptions, data quality, parameter selection, and model structure. Simulation results should be treated as research and decision-support evidence, not as automatic proof of real-world causality. Medical, biological, engineering, and safety-critical conclusions require appropriate domain-specific validation.

専門分野

• ☆ Aviation
• ☆ Industrial Engineering
• ☆ Environmental Science
• ☆ Ecology
• ☆ Cryogenics
• ☆ Industrial Design

サービス

• Writing
• Research
• Consulting
• Data & AI

Writing Medical Writing

Research Feasibility Study, Gap Analysis, Gray Literature Search, Scientific and Technical Research

Consulting Operations Consulting, Scientific and Technical Consulting

Data & AI Predictive Modeling, Statistical Analysis, Algorithm Design-Non ML, Algorithm Design-ML, Data Visualization

職務経験

Independent Researcher

Independent Research Program

2月 2026 - 現在

学歴

Associate Degree (Hotel Management)

Jeongeup

3月 2004 - 3月 2008

Hotel Management

Jeonbuk Science College - korea

3月 2004 - 2月 2008

認定資格

• Wordpress Level 2

  Korea Chamber of Commerce and Industry

  1月 1997 - 現在

出版物

SOFTWARE

Min Gi Kim (2026). OrganOS RAG Candidate Compression .

Min Gi Kim (2026). OrganOS Thermal Memory Architecture .

Min Gi Kim (2026). OrganOS Thermal-Echo Drug Discovery Benchmark .

Min Gi Kim (2026). OrganOS-ChEMBL-Benchmark .