Senior AI Security Engineer

Quarks Technosoft Pvt. Ltd.; Pune District; Sur place; INR 1 200 000 - 2 300 000; Plein temps

Résumé du poste

Quarks Technosoft Pvt. Ltd. in Pune seeks a seasoned security engineer with a focus on AI safety and cloud security. The role emphasizes hands-on engineering practice and practical security implementation for AI systems.

You will demonstrate red-teaming AI systems, prompt-injection countermeasures, and reproducible reporting. Proficiency in Python, CI/CD, IaC, and Azure, plus embeddings and vector databases knowledge, is required.

Qualifications

• Bachelor’s degree or equivalent experience; practical engineering experience counts.
• At least 6 years in security engineering, app security, offensive security, detection engineering or cloud security.
• 1–2 years hands-on work securing production generative AI/LLM systems.
• Experience red-teaming AI systems against prompt injection, jailbreaks and model abuse.
• Strong Python with ability to build test harnesses and tooling to production standards.
• Solid cloud security experience on a major hyperscaler; Azure preferred.
• Understanding of LLM architectures: embeddings, vector databases, RAG, agent frameworks.
• Familiarity with OWASP Top 10 for LLM apps and MITRE ATLAS guidance.
• CI/CD and IaC fluency; embed security checks in delivery pipelines.
• Experience with current AI development tooling to build tooling and evaluation harnesses.

Responsabilités

• Develop and implement security controls for AI systems.
• Conduct red-team exercises focused on prompt injection, jailbreaks and abuse.
• Build reproducible, actionable security reports for engineering teams.

Connaissances

Python

Security engineering

Offensive security

Cloud security

LLM security

Red-teaming AI systems

Prompt injection

Model abuse mitigation

Automation tooling

Agent frameworks

Formation

Bachelor’s degree

Outils

Azure

CI/CD tooling

Infrastructure as Code

AI development tooling

Agent frameworks

Description du poste

• Bachelor’sdegree in any discipline, or equivalent professional experience; demonstratedengineering practice is the primary criterion.
• Minimum6 years in security engineering, application security, offensive security,detection engineering or cloud security.
• Minimum1–2 years of hands-on work securing or testing production generative AI, LLM oragentic systems.
• Demonstratedexperience red-teaming AI systems against prompt injection, jailbreaks,excessive agency and model abuse — candidates should be able to walk throughspecific findings they produced.
• StrongPython, with the ability to build test harnesses, automation and tooling to aproduction standard.
• Solidcloud security engineering experience on a major hyperscaler; Azure preferred.
• Practicalunderstanding of LLM application architecture: embeddings, vector databases,RAG, agent frameworks, tool and function calling, and the MCP-style integrationpattern.
• Workingknowledge of OWASP Top 10 for LLM Applications, OWASP agentic security guidanceand MITRE ATLAS.
• CI/CDand infrastructure-as-code fluency; comfortable embedding security checks intodelivery pipelines.
• Practicaluse of current AI development tooling — agent-building platforms, AI-assistedcoding environments and locally hosted models — to build tooling and evaluationharnesses at speed.
• Clearwritten reporting — findings must be reproducible and actionable by thereceiving team.