Lead Adversarial AI Security Engineer

Finastra; Atlanta (GA); Hybrid; USD 120,000 - 180,000; Full time

Benefits offered by this job

Unlimited vacation

Hybrid work

Health insurance

Retirement plan

Wellbeing program

Volunteer time off

Job summary

Finastra, a global financial services software leader, seeks an experienced security researcher to design and execute adversarial AI attacks, testing defenses, detection, and response capabilities. You will simulate red team operations on AI/ML systems and produce actionable PoCs.

You will collaborate with cyber defense teams, prepare detailed reports, and develop tooling to automate assessments while exploring emerging security technologies and vendor solutions.

Qualifications

• Proven experience with offensive security/research.
• Demonstrated adversarial AI; Python; replay/fuzzing automation; Security testing standard or Framework such as the MITRE ATLAST Framework
• Adversarial mindset and critical thinking skills

Responsibilities

• Lead the planning, design and implementation of assessments and plans, with a focus on AI/ML systems. Continuously test LLM/RAG/agent systems for adversarial risks and automate in CI.
• Perform full scope adversarial simulation/red team assessments such as design and run adversarial suites (prompt injection, jailbreaks, retrieval poisoning), test model extraction, membership inference, and agent tool-use abuse, deliver PoCs and fixes.
• Work with the cyber defense Teams to develop new and improved ways to defend the organization
• Prepare reports containing attack paths, findings/vulnerability information, and mitigation options and recommendations
• Provide support during ongoing assessments by writing tools and setting up infrastructure
• Creating automation for specific tasks and operations
• Research, investigate and evaluate emerging technologies, industry trends and best practices, and vendor solutions to be able to recommend and assess various security products

Skills

Offensive security

Adversarial AI

Python

Automation (replay/fuzzing)

MITRE ATLAST Framework

Job description

Finastra, a global financial services software leader, seeks an experienced security researcher to design and execute adversarial AI attacks, testing defenses, detection, and response capabilities. You will simulate red team operations on AI/ML systems and produce actionable PoCs.

You will collaborate with cyber defense teams, prepare detailed reports, and develop tooling to automate assessments while exploring emerging security technologies and vendor solutions.