Principal Security Operation Engineer (Red Team)

About Us

• Responsible for developing and executing penetration testing, red-blue confrontation, and practical attack and defense drills that simulate real attack scenarios, identifying potential security risks in enterprise networks, applications, cloud environments, work networks, and core business systems.
• Lead or participate in red-blue confrontation exercises to evaluate the defense team's ability in attack detection, alarm analysis, traceability analysis, emergency response, and recovery.
• Based on the real attack chain design exercise scenario, covering extranet breakthrough, web vulnerability exploitation, phishing entrance, privilege escalation, lateral movement, Data Discovery, privilege maintenance, and defense bypass stages.
• Combining the attack review results, promote the continuous optimization of security detection rules, response processes, asset governance, and security base lines.

• Identify enterprise network exposure, internet assets, cloud assets, APIs, supply chain components, and third-party access risks, evaluate attack paths, and provide mitigation recommendations.
• Monitor and collect threat intelligence, track vulnerability exploitation trends, APT attack methods, red team toolchain changes, and apply them to enterprise attack and defense exercises.
• Combining business scenarios to model attack paths and discover feasible attack chains from external exposure surfaces to core assets.
• Tracking AI-related security risks, including security issues in large-scale model applications, RAG systems, Agent systems, plug-in/tool calls, MCP services, AI code generation, and automated workflows.

• Responsible for the security evaluation of AI applications, intelligent agent systems, RAG Knowledge Base, AI Agent toolchain and model services within the enterprise.
• Research and verify large-scale model-related attack techniques, including Prompt Injection, Jailbreak, Indirect Prompt Injection, data leakage, unauthorized tool invocation, security risks caused by model illusions, RAG poisoning, vector library pollution, sensitive information leakage, etc.
• Design AI red team test cases and evaluation framework, and conduct security verification on model input and output, context isolation, permission control, tool call chain, and data access boundary.
• Participate in the construction of AI security protection plan, including prompt word security policy, content security detection, tool call permission constraints, sensitive data desensitization, audit tracking, Agent sandbox isolation and security evaluation benchmark construction.
• Explore the application of AI in red team automation, vulnerability analysis, attack path planning, PoC verification, and report generation, and promote the platformization, automation, and intelligence of attack and defense capabilities.

• Develop and optimize Red Team-specific tools and scripts for vulnerability mining, information collection, privilege escalation, lateral movement, credential analysis, traffic disguise, defense bypass, and automated report generation.
• Study and validate new attack techniques, and simulate real threats in combination with enterprise business scenarios.
• Build or participate in the construction of automated security evaluation platform, integrated vulnerability scanning, audio fingerprint recognition, asset mapping, PoC verification, attack path analysis, AI Agent arrangement and other capabilities.
• Combining LLM/Agent technology to explore automated penetration testing, intelligent vulnerability verification, code security auditing, and red team task scheduling.

• Conduct security evaluations on critical business systems, internal networks, cloud environments, work end points, API services, and AI applications, and output detailed technical reports, attack paths, impact analysis, and repair recommendations.
• Output AI security evaluation reports for AI applications, including attack examples, risk levels, exploitable paths, data leakage risks, permission boundary issues, and governance recommendations.
• Assist in improving enterprise security protection mechanisms, promote optimization of WAF, EDR, SIEM, NDR, HIDS, zero trust, identity permissions, and log auditing capabilities.
• Transform attack and defense discovery into security detection rules, base line specifications, develop security requirements, and continuous governance mechanisms.

• Collaborate with the blue team, security operation, infrastructure, R & D, algorithm, data, and business teams to complete attack review, vulnerability repair, detection rule optimization, and security capability building.
• Provide security support to other departments of the enterprise, including emergency response drills, development security consulting, AI application pre-launch security review, and security training.
• Participate in the security design review of AI applications and security products, and promote the advance of security capabilities in R & D, testing, and Pushonline.

• Proficient in basic knowledge of cyber security, including TCP/IP protocol, network architecture, identity authentication, access control, principles and configurations of common security devices.
• Proficient in common attack techniques and red team toolchains, such as Sliver, Cobalt Strike, NPS, Burp Suite, Metasploit, Nmap, Masscan, Frida, Impacket, etc.
• Familiar with mainstream operating systems Windows, Linux, macOS security mechanism, log system, permission model and common use.
• Familiar with common web frameworks, API architectures, microservice structures, containers, and security risks in Kubernetes environments.

• Proficient in penetration testing and red team processes, including information collection, vulnerability scanning, vulnerability exploitation, intranet penetration, privilege escalation, lateral movement, privilege maintenance, Data Discovery, and trace cleaning.
• Familiar with the working principles and adversarial methods of enterprise-level security products, including WAF, EDR, SIEM, NDR, HIDS, zero-trust gateway, bastion host, and identity authentication system.
• Possess independent vulnerability analysis and PoC writing capabilities, able to reproduce, verify, and assess the impact of public vulnerabilities and 0day/1day risks.
• Proficient in one or more programming/scripting languages, such as Python, Go, Bash, PowerShell, JavaScript, etc., with experience in tool development and automation platform construction.

• Familiar with the basic architecture of large-scale model applications, familiar with technical forms such as Prompt, RAG, Embedding, vector databases, Agent, Function Calling, MCP, and plugin systems.
• Understand or practice AI security testing methods, including Prompt Injection, Jailbreak, RAG data poisoning, sensitive information leakage, unauthorized tool invocation, model output security, Agent permission escape, etc.
• Able to design security test cases for AI applications, evaluate model input/output, context isolation, data boundaries, permission control, and tool invocation chain risks.
• Experience in using AI to assist security work, including vulnerability analysis, code auditing, intelligence analysis, attack path planning, report generation, or automated testing.
• Familiar with AI application security governance ideas, including model call auditing, prompt word security, sensitive data protection, content security detection, permission minimization, and sandbox isolation.

• More than 5 years of experience in red team, penetration tes