Secure AI Systems Engineer

Bright Vision Technologies is a technology consulting and software development company delivering cloud, AI, data, and enterprise solutions across the United States. The Secure AI Systems Engineer will lead the design and implementation of security controls, threat models, and incident response capabilities for AI and machine learning systems, while addressing risks involving LLMs, model APIs, training data pipelines, and AI-powered applications.

Responsibilities:

• Lead the design and implementation of security controls, threat models, and incident response capabilities specifically tailored to AI and machine learning systems
• Addresses the unique security challenges posed by LLMs, model APIs, training data pipelines, and AI-powered applications, including prompt injection, model abuse, data exfiltration, and supply chain risks
• Designing defenses for both AI-powered applications and the AI infrastructure that supports them

Requirements:

• Bachelor's or Master's degree in Computer Science, Cybersecurity, or a related discipline
• Six or more years of security engineering experience, including significant work on AI or ML systems
• Strong understanding of LLM internals, modern AI architectures, and common failure modes
• Hands-on experience designing security controls for AI-powered applications
• Deep knowledge of application security, identity, and cryptography fundamentals
• Experience with threat modeling and security architecture review processes
• Familiarity with adversarial ML, prompt injection, and model abuse research
• Proficiency in Python and at least one systems language
• Strong understanding of cloud security and modern infrastructure controls
• Excellent written and verbal communication skills
• Publications, talks, or CTF participation in AI security topics
• Experience with red-teaming LLM-based products
• Familiarity with privacy-preserving ML techniques such as differential privacy
• Exposure to regulated industries with strict data handling requirements
• Open-source contributions to AI security tooling