Security Engineer

About NeuralTrust

AI agents are moving into production fast, and most companies have no way to see or control what those agents do. NeuralTrust is the security layer for AI agents. We’re a Barcelona-based company backed by a €20M seed round.

Teams use our platform to secure their agents at runtime, govern how agents reach models and tools through our Agent Gateway, and test them against real-world attacks before attackers get there first.

Gartner recognizes us in its Market Guides for Guardian Agents and for AI Gateways. Today we protect production AI workloads for customers across Europe and the US.

The role

You’ll join the engineering team as a hands-on Security Engineer and build security into the core of our platform.

You’ll design, build and ship the protections our customers rely on: guardrails, policy enforcement, detection logic and secure-by-default features across our gateway and runtime security products.

You’ll work closely with our AI security researchers, who find new attack techniques against LLMs and agents. Your job is to turn those findings into defenses that are reliable, fast and ready for production.

What you’ll do

• Design and build security features for our Agent Gateway (TrustGate) and Agent Runtime Security (TrustGuard). Examples: prompt injection defenses, data loss prevention, tool and MCP access controls, rate limiting and policy enforcement.
• Turn research on emerging threats into production defenses. These include jailbreaks, indirect prompt injection, tool misuse and data exfiltration.
• Build and harden the authentication, authorization and identity flows for AI agents and the tools they call.
• Run threat modeling and secure design reviews for new features. Partner with engineers to fix issues early.
• Improve detection quality with the ML team, keeping both false positives and latency low.
• Own security across the SDLC for the products you touch: code review, dependency and container scanning, secrets management, and secure CI/CD.
• Help respond to security issues reported by customers or researchers.

What you bring

• 3–5 years of experience in software engineering, product security or application security.
• Strong coding skills in at least one backend language. Go or Python is ideal.
• A solid grasp of web and API security: OWASP Top 10, authN/authZ (OAuth, OIDC, JWT), input validation, and secure API design.
• Experience building or securing distributed systems in the cloud (AWS, GCP or Azure), with Docker and Kubernetes.
• You enjoy shipping. You’d rather build the control than write the report about it.
• Clear communication and the ability to explain security trade-offs to engineers and customers.
• Professional working proficiency in English.

Nice to have

• Hands-on experience with LLMs, AI agents or agent frameworks (LangChain, OpenAI Agents SDK, MCP, A2A).
• Knowledge of the OWASP Top 10 for LLM Applications or MITRE ATLAS.
• Experience with AI or MCP gateways, reverse proxies or high-throughput, low-latency services.
• Background in detection engineering, WAFs or runtime protection.
• Contributions to open-source security projects, CVEs or bug bounty findings.
• Spanish is a plus.

Why NeuralTrust

• Work on one of the fastest-moving problems in security: protecting AI agents in production.
• Real ownership in a small, senior team where your work ships to customers quickly.
• Remote-friendly across Spain and the EU, with a Barcelona HQ for team time.
• Competitive salary