Marshall Goodman

MARSHALL H. GOODMAN

Business Analyst, AI Security | StrategicPromptArchitect@gmail.com | StrategicPromptArchitect.com

PROFESSIONAL SUMMARY

Seasoned IT professional with 40+ years of progressive experience spanning mainframe systems to cutting-edge AI security. Founder of StrategicPromptArchitect, specializing in AI security consulting with deep expertise in malicious prompt detection, workflow security, and business-technical requirements analysis. Proven ability to identify emerging security vulnerabilities in AI systems and develop practical solutions. Combines technical depth with strategic business analysis to bridge the gap between AI capabilities and enterprise security requirements.

CORE COMPETENCIES

• AI Security Architecture: Prompt injection detection, malicious prompt intent (MPI) analysis, indirect prompt injection, context poisoning, workflow security frameworks

• AI Development & Integration: ChatGPT API, Claude API, custom initialization prompts, meta-prompting systems, persona-based AI optimization

• Security Tool Development: Browser extensions (JavaScript), Python security modules, MCP server prototypes, workflow platform security (n8n, Node-RED)

• Business Analysis: Requirements elicitation, strategic questioning methodology, AI risk assessment, technology-business alignment

• Database & Systems Design: MySQL, SQL optimization, entity-relationship modeling, multi-tier architecture, COBOL, Visual Basic, JavaScript, Python

KEY AI SECURITY PROJECTS

Sentinel.AI - Chrome Extension for Prompt Injection Protection 2024-Present | Open-source browser security extension • Designed and developed real-time prompt injection detection system using JavaScript and pattern matching algorithms

• Implemented 13+ threat categories including jailbreaks, data exfiltration attempts, instruction overrides, and context poisoning

• Created drag-and-drop prompt deck scanner for batch security analysis of AI workflow templates

• Achieved local-only processing architecture with zero external data transmission, ensuring complete user privacy

MalPromptSentinel - Multi-Platform Security Framework 2024-Present | Cross-platform malicious prompt detection • Developed Chrome extension variant with advanced zip file scanning capabilities for workflow template security

• Created Anthropic Skills prototype for Claude AI integration, enabling context-aware security validation in agent workflows

• Researched and documented attack vectors for multi-node AI workflows including MCP server vulnerabilities

• Pioneered workflow-level threat assessment methodology for enterprise AI automation platforms

AI Security Research & Thought Leadership 2023-Present | Published technical analysis and security frameworks • Authored comprehensive articles on AI security attack surfaces, workflow vulnerabilities, and defensive strategies (StrategicPromptArchitect.com)

• Published analysis of Anthropic Skills security implications and emerging attack surface complexities

• Developed meta-prompting frameworks for optimized AI interactions and hallucination reduction

• Created specialized persona systems for coordinated AI expert workflows and domain-specific problem-solving