Senior AI Threat Modeling & Security Architect

emagine Polska; Województwo pomorskie; On-site; PLN 120,000 - 180,000; Part time

Job summary

emagine Polska is seeking a Threat Modeling Specialist to advance security posture by identifying AI-related threats within the Mythos framework. You will conduct threat assessments, map assets and attack surfaces, and collaborate with development and security teams to define mitigations.

The role is part-time (20% per month) with occasional office visits to Gdańsk, Warsaw, or Łódź, requiring hands-on experience with STRIDE, attack trees, kill chains, and OWASP LLM Top 10 familiarity.

Qualifications

• Experience in IT security architecture and threat modeling.
• Practical use of STRIDE, attack trees and kill chains.
• Ability to map assets, trust boundaries and attack surfaces.
• Experience with AI threats related to agents, ML and LLM.

Responsibilities

• Conduct application-level threat assessments and map Mythos threat scenarios.
• Facilitate threat modeling workshops with application and platform teams.
• Identify assets, trust boundaries, attack surfaces, threats and associated risks.
• Develop and maintain threat models utilizing approved tools and methodologies.
• Review solution designs and architecture to ensure comprehensive threat coverage.
• Collaborate with development and security teams to define and implement risk mitigation strategies.
• Track and report on identified threats, risks and their remediation status.

Skills

Threat modeling

STRIDE

Attack trees

Kill chains

AI/LLM threats

IT security arch

Asset mapping

Job description

emagine Polska is seeking a Threat Modeling Specialist to advance security posture by identifying AI-related threats within the Mythos framework. You will conduct threat assessments, map assets and attack surfaces, and collaborate with development and security teams to define mitigations.

The role is part-time (20% per month) with occasional office visits to Gdańsk, Warsaw, or Łódź, requiring hands-on experience with STRIDE, attack trees, kill chains, and OWASP LLM Top 10 familiarity.