Experience & Qualifications
10+ years of cybersecurity experience, including 2+ years of practical AI security experience.
Strong technical engineering background with the ability to assess, design, implement, and validate security controls for enterprise AI environments.
Deep understanding of AI architectures, including LLMs, AI gateways, agentic AI, MCP, and associated security risks.
Security-focused resource rather than a general AI/ML engineer or high-level security strategist.
Able to operate independently, identify gaps, make informed recommendations, and execute technical work.
Key Responsibilities
AI Gateway Evaluation: Assess AI gateway technologies, evaluate security capabilities, and help select and implement the appropriate solution.
Agentic AI & MCP Security: Evaluate MCP implementations, agent permissions, authentication/authorization, data access, scope limitations, and associated security risks.
AI Architecture Security Reviews: Review proposed AI architectures and data flow diagrams to identify missing components, security gaps, trust boundaries, and necessary controls.
Security Controls Implementation: Recommend, configure, and implement technical security controls for enterprise AI applications and agentic workflows.
Microsoft AI Security: Experience with Microsoft's AI ecosystem, particularly AI agents, gateways, MCP integration, and associated security configuration and policy enforcement.
Technical Advisory: Provide practical, actionable security recommendations and collaborate with engineering and architecture teams to implement them.
Engagement & Commercial Requirements
Full-time staff augmentation, initially targeting a 6-month contract rather than T&M.
Requesting pricing and cost breakdown for a US-based resource.
Open to an India-based resource if technical qualifications and experience are a strong match; comparative pricing requested.
Houston-based or hybrid resource (1–2 days/week onsite) is an option, but not a strict requirement, person can be 100% remote - customer is flexible for the right resource.
