Cybersecurity Experts for AI Training

Company

Gramian Consultancy

Salary / Compensation

• Compensation not specified

Overview / Job Description

Gramian Consultancy is a boutique consulting firm specializing in IT professional services and engineering talent solutions. The company helps organizations build high-performing technical teams by connecting them with highly qualified professionals.

This role focuses on developing advanced evaluation environments for frontier AI systems. The Cybersecurity Expert will design realistic security-focused coding challenges, adversarial testing environments, and vulnerability evaluation benchmarks that assess how AI models understand cybersecurity concepts, exploit chains, software vulnerabilities, and secure coding practices.

The position combines offensive security, application security, software engineering, and AI evaluation to create reproducible security testing environments inspired by real-world attack scenarios and vulnerability patterns.

Responsibilities

Security Scenario Design & Development

• Design vulnerable multi-component applications for AI evaluation
• Build realistic security challenge environments across multiple programming languages
• Create scenarios involving web applications, APIs, authentication systems, and distributed architectures
• Develop reproducible environments that simulate real-world attack surfaces
• Design evaluation tasks that test AI reasoning in cybersecurity contexts

Vulnerability Research & Exploit Development

• Develop exploit chains involving multiple vulnerability classes
• Create adversarial security scenarios based on real-world attack techniques
• Model vulnerabilities inspired by:
  • CVEs
  • Bug bounty discoveries
  • Production security incidents
• Design hidden attack paths and complex exploitation workflows
• Simulate advanced threat scenarios for AI testing

AI Security Evaluation

• Review and analyze AI-generated outputs
• Identify weaknesses in AI security reasoning and vulnerability detection
• Evaluate remediation recommendations generated by AI systems
• Assess exploit identification and secure coding capabilities
• Create benchmarks for AI security performance testing

Testing Environment Development

• Build deterministic evaluation environments using Docker
• Develop automated validation and testing frameworks
• Create security-focused test cases and verification logic
• Ensure environments remain scalable and reproducible
• Implement mechanisms that prevent benchmark contamination

Adversarial Testing & Benchmarking

• Develop misleading documentation and edge-case scenarios
• Create obfuscated code samples and hidden vulnerabilities
• Design adversarial test cases that challenge AI systems
• Evaluate model robustness against complex security challenges
• Improve benchmark reliability and evaluation quality

Cross-Functional Collaboration

• Work with AI evaluation and benchmarking teams
• Collaborate with software engineers and researchers
• Support automated testing and validation initiatives
• Contribute to security-focused AI evaluation frameworks
• Participate in technical discussions and review processes

Requirements

Experience

• 4+ years of experience in:
  • Cybersecurity
  • Application Security
  • Vulnerability Research
  • Offensive Security
• Experience with vulnerability discovery and analysis
• Experience conducting secure code reviews
• Experience with exploit development or patch validation
• Experience working in security testing environments
• Experience building or analyzing complex software systems

Security Expertise

• Web Application Security
• Authentication & Authorization Systems
• Session Management Security
• OAuth Security
• JWT Security
• SSRF Vulnerabilities
• Injection Attacks
• Access Control Vulnerabilities
• Privilege Escalation
• Filesystem Security
• Cryptographic Vulnerabilities

Programming Skills

• Strong coding ability in at least two of the following:
  • Go
  • Python
  • Node.js
  • Rust
• Experience building and analyzing production software systems
• Ability to develop realistic vulnerable applications

Technical Skills

• Docker
• Containerized Environments
• Linux Systems
• Security Testing Frameworks
• Automated Validation Tools
• Vulnerability Assessment
• Security Benchmark Development
• AI Evaluation Environments

Security Tools

• SAST (Static Application Security Testing)
• IAST (Interactive Application Security Testing)
• Fuzzing Tools
• Security Testing Frameworks
• Vulnerability Scanners
• Secure Code Analysis Tools

Candidate Profile

You SHOULD Apply If

• You have extensive hands-on cybersecurity experience
• You enjoy vulnerability research and exploit development
• You have strong software engineering capabilities
• You are interested in AI safety, evaluation, and benchmarking
• You enjoy creating realistic attack scenarios and security challenges
• You have experience with offensive security methodologies
• You are comfortable working with ambiguity and advanced technical problems
• You enjoy combining security expertise with emerging AI technologies

You Should NOT Apply If

• You have limited hands-on security experience
• You lack programming proficiency in modern development languages
• You are unfamiliar with vulnerability research or exploit analysis
• You prefer compliance-focused security work over technical security research
• You are uncomfortable building and testing intentionally vulnerable systems
• You have limited experience with Linux and containerized environments

Other Information

Additional Benefits / Highlights

• Opportunity to contribute to frontier AI evaluation systems
• Work on cutting-edge AI security and benchmarking initiatives
• Exposure to advanced cybersecurity research environments
• Highly technical and hands-on role
• Opportunity to influence AI safety and security evaluation methodologies
• Work with cross-functional engineering and research teams
• Challenging and intellectually stimulating projects
• Remote contractor opportunity

Preferred Qualifications

• Bug bounty experience
• CTF (Capture The Flag) participation
• Red team experience
• CVE research contributions
• Security research publications
• Advanced exploit development experience
• AI security or AI safety experience

Work Schedule

• 8 hours per day
• Must maintain at least 4 hours overlap with PST timezone
• Remote work arrangement

Contract Details

• Contractor assignment
• Minimum duration of 4+ weeks
• No medical or paid leave benefits

Eligible Locations

• Nigeria
• Ghana
• Kenya
• Egypt
• Turkey
• India
• Pakistan
• Bangladesh
• Indonesia
• Vietnam
• Brazil
• Colombia

Interview Process

• Two technical interview rounds