AI Security Engineer — On-Device & Hardware-Backed (Hybrid)

XFA; Antwerpen; Hybride; EUR 70 000 - 110 000; Plein temps

Avantages offerts par ce poste

Company insurance

Meal vouchers

Commute coverage

13th month

Résumé du poste

XFA in Antwerp, Belgium is hiring an AI Engineer to join a hybrid team at the intersection of AI and cybersecurity. You will help secure AI agents across devices and cloud components, implementing identity, authentication, and authorization with hardware-backed security.

You will work on threat modeling, cryptographic protections, and privacy-preserving AI pipelines, collaborating with R&D to push secure, production-ready AI tooling and MCP integration.

Qualifications

• Several years shipping production software in Golang, Python, Typescript, or systems language (Rust/C++).
• Strong understanding of LLMs and AI agents including tools, prompts, and orchestration.
• Experience with LLM tooling or MCP protocol is a plus.
• Solid security engineering background: threat modeling, cryptography, auth protocols.
• Knowledge of hardware-backed security (Secure Enclave, TPMs) and confidential computing.
• Familiarity with OS internals and endpoint security basics.
• Experience with policy-as-code (e.g., Cedar, OPA) and identity/workload security concepts.
• Fluent in English; additional European languages are a plus.

Responsabilités

• Design and build identity and authentication for AI agents with hardware-backed keys.
• Develop policy-based authorization so authority narrows at each handoff.
• Build on-device enforcement for agent actions with low latency.
• Detect threats like unknown agents, behavior drift, or prompt tampering.
• Design privacy-preserving AI pipelines with on-device processing and remote attestation when needed.
• Apply LLMs where they add value, balancing prompts, retrieval, and local models.
• Create evaluation pipelines, attack simulations, and automated tests for reliability.
• Stay updated on AI agents, security standards, and emerging threats.

Connaissances

Golang

Python

Typescript

Rust/C++

LLMs/AI agents

Security engineering

Threat modeling

OS internals

Policy as code

Privacy-preserving

English proficiency

Outils

MCP

Cedar

OPA

SPIFFE/SPIRE

OAuth 2.0

OpenID Connect

JWT

Description du poste

XFA in Antwerp, Belgium is hiring an AI Engineer to join a hybrid team at the intersection of AI and cybersecurity. You will help secure AI agents across devices and cloud components, implementing identity, authentication, and authorization with hardware-backed security.

You will work on threat modeling, cryptographic protections, and privacy-preserving AI pipelines, collaborating with R&D to push secure, production-ready AI tooling and MCP integration.