Research engineer position on secure agentic ai systems
Pubblicato il 21-08-2026 - AI4I in Italia
Research Engineer Position on Secure Agentic AI Systems AI4 I invites applications for the Research Engineer role to join our founding team and drive secure agentic AI systems. Diagnostic Deadline: December 14, 2025, 11:59 PM CEST. The Role As a research engineer you will design and implement a scalable end‐to‐end security platform enabling secure AI deployment at scale. Your contributions will translate theoretical security concepts into production‐ready systems, forming foundational infrastructure for AI red teaming, secure agent execution environments, verification protocols, and continuous monitoring frameworks that protect AI systems during runtime and safeguard data throughout processing, storage, and transfer. Key Responsibilities Design and implement scalable security platforms for AI agents and large language model workloads, including secure execution environments and runtime protection mechanisms. Conduct proactive red team exercises simulating external adversaries and insider threats to identify and remediate vulnerabilities in agentic AI solutions.
Develop and deploy defenses against LLM‐specific threats including prompt injection, task hijacking, model extraction, and data leakage attacks. Build security validation frameworks and compliance certification tools to support secure system deployments for internal teams and pilot partners. Collaborate with research teams to translate novel security findings into production‐ready implementations and open‐source security tools. Minimum Qualifications Master's or Ph D in Computer Science, Engineering, or a related field with focus on security, systems, or machine learning. Hands‐on experience with modern ML frameworks including Py Torch, Hugging Face, JAX, or Tensor Flow for deploying and securing AI workloads. Preferred Qualifications Experience conducting penetration testing, vulnerability assessments, security architecture reviews, or threat modeling for complex systems. Expertise with trusted execution env