Forward Deployed AI Engineer for Cyber Security

Description

You will combine applied LLM engineering with enterprise security, embedded in the application security, SOC, threat intelligence and vulnerability management teams.

Your work is to build, deploy and scale production AI systems, LLM pipelines and autonomous agent workflows that tackle hard security problems.

This is not API wiring or chat tool use. You will design and ship robust, tested, tool-using LLM systems and automated security capabilities into live production at a global financial institution.

Threats here are active and the work goes straight to production. Teams are engineering-led and small, with high autonomy and minimal governance overhead. Executive sponsorship helps you move fast, and the frameworks you build may shape how the industry responds.

This is one requisition for the Cyber AI Engineering team, open at Assistant Vice President (AVP, C12) or Vice President (VP, C13) level. Placement depends on technical depth, engineering leadership and domain expertise.

Responsibilities

• Build, deploy and run LLM-based security applications, tool-using agents and retrieval-augmented generation (RAG) pipelines for tasks such as automated vulnerability remediation, SOC alert triage, SAST and DAST triage, and detection engineering.
• Work alongside AppSec engineers, SOC analysts, incident responders and penetration testers to find where their work slows down, then turn those needs into working AI software.
• Design evaluated, tool-using agents that call enterprise APIs, security scanners, SIEM and SOAR platforms, code repositories and containerized infrastructure in a secure way.
• Set up evaluation frameworks that measure latency, accuracy, faithfulness, hallucination and tool-call precision, and add input and output guardrails for predictable behavior.
• Write clean, modular Python, package services as containers, and release them through automated CI/CD pipelines that meet enterprise security and architecture standards.
• Defend against AI-specific attacks such as prompt injection, model abuse, data leakage and unsafe tool execution, using secure coding practices from the start.
• Build reusable AI components, evaluation patterns and integration libraries that teams across the CISO engineering organization can adopt.

Requirements

• You have built LLM-powered systems yourself, including prompt engineering, RAG, structured outputs and tool or function calling, not only used commercial chat assistants.
• You have foundational knowledge of application security (OWASP Top 10, SAST, DAST, secure code review) or security operations (SOC alert triage, SIEM workflows, incident analysis).
• Nice-to-have: 4-7+ years of software engineering experience, strong Python, and a record of shipping production services or pipelines such as REST APIs, microservices, Docker/Kubernetes and CI/CD.
• Nice-to-have: hands-on work with AppSec tools such as Semgrep, CodeQL, Snyk or Burp Suite, or enterprise SIEM platforms such as Splunk, Microsoft Sentinel, Elastic or Google Chronicle.
• Nice-to-have: familiarity with detection engineering, the MITRE ATT&CK framework, and common AI attack surfaces such as prompt injection, jailbreaking and the OWASP Top 10 for LLMs.
• You have architected, built and evaluated production tool-using agent systems, including multi-step orchestration, RAG, guardrails, and evaluation of accuracy, latency and reliability.
• You have strong depth in either offensive security or AppSec, or in detection and SOC, plus working fluency and cross-functional understanding of the other.
• Nice-to-have: 8-10+ years of software engineering experience, deep Python expertise, and a record of architecting, shipping and owning production systems end to end.
• Nice-to-have: foundational security certifications such as OSCP, GWAPT, GPEN or GCIH.
• Nice-to-have: strong awareness of AI security risks (prompt injection, model poisoning, model abuse, data leakage), hands-on fine-tuning or adaptation experience, and SOAR workflow development.

Conditions

Full-time role based in Hungary, under a hybrid model with up to 2 days a week working from home.

Salary range: 12,211,560 Ft to 28,767,800 Ft.

The company is an equal opportunity employer, and its EEO policy statement and Know Your Rights poster apply.

Benefits

• Competitive compensation package.
• Cafeteria program.
• Home office allowance for hybrid work.
• Paid parental leave for maternity and paternity.
• Private medical care and onsite medical rooms at the offices.
• Pension contribution to a voluntary pension fund.
• Group life insurance.
• Employee assistance service.
• Learning and development programs, online course libraries and upskilling platforms such as Udemy and Degreed.
• Flexible work arrangements.
• Career progression across geographies and business lines.
• Employee communities for networking.