Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I’m an AI-focused Red Teaming and automation specialist who helps enterprises evaluate and harden large language model systems. I build and run security assessments that stress AI applications—especially RAG pipelines—against prompt injection, systemic vulnerabilities, and risky agent behaviors. I also specialize in end-to-end AI automation workflows on isolated Linux architectures, including automated vulnerability scanning and structured model security auditing. With hands-on experience in tools like Microsoft PyRIT, Garak, and RAG dataset evaluation, I support organizations with practical, enterprise-ready security evaluation processes.…I’m an AI-focused Red Teaming and automation specialist who helps enterprises evaluate and harden large language model systems. I build and run security assessments that stress AI applications—especially RAG pipelines—against prompt injection, systemic vulnerabilities, and risky agent behaviors. I also specialize in end-to-end AI automation workflows on isolated Linux architectures, including automated vulnerability scanning and structured model security auditing. With hands-on experience in tools like Microsoft PyRIT, Garak, and RAG dataset evaluation, I support organizations with practical, enterprise-ready security evaluation processes.

Mf Mashfi

AI Trainer, AI Prompt Engineer, AI Ethics Consultant, +2





I’m an AI-focused Red Teaming and automation specialist who helps enterprises evaluate and harden large language model systems. I build and run security assessments that stress AI applications—especially RAG pipelines—against prompt injection, systemic vulnerabilities, and risky agent behaviors. I also specialize in end-to-end AI automation workflows on isolated Linux architectures, including automated vulnerability scanning and structured model security auditing. With hands-on experience in tools like Microsoft PyRIT, Garak, and RAG dataset evaluation, I support organizations with practical, enterprise-ready security evaluation processes.…I’m an AI-focused Red Teaming and automation specialist who helps enterprises evaluate and harden large language model systems. I build and run security assessments that stress AI applications—especially RAG pipelines—against prompt injection, systemic vulnerabilities, and risky agent behaviors. I also specialize in end-to-end AI automation workflows on isolated Linux architectures, including automated vulnerability scanning and structured model security auditing. With hands-on experience in tools like Microsoft PyRIT, Garak, and RAG dataset evaluation, I support organizations with practical, enterprise-ready security evaluation processes.

Available to hire

I’m an AI-focused Red Teaming and automation specialist who helps enterprises evaluate and harden large language model systems. I build and run security assessments that stress AI applications—especially RAG pipelines—against prompt injection, systemic vulnerabilities, and risky agent behaviors.

I also specialize in end-to-end AI automation workflows on isolated Linux architectures, including automated vulnerability scanning and structured model security auditing. With hands-on experience in tools like Microsoft PyRIT, Garak, and RAG dataset evaluation, I support organizations with practical, enterprise-ready security evaluation processes.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Work Experience

Ai red teaming at Scaleflow ai

October 14, 2026 - Present

AI Red Teaming & Full-Stack Automation Specialist (Scale Flow AI founder) at Scale Flow AI

January 1, 2020 - Present

Deployed Microsoft PyRIT and Garak on Ubuntu Linux terminals to safely execute bulk prompt-injection vulnerability testing. Simulated stateful RAG proxy architectures to intercept and mitigate secure-vector database data leaks. Evaluated model resilience against indirect prompt injection and agentic over-agency behaviors, including security auditing approaches for RAG and system-level weaknesses.