Security Architect

Role overview

Responsibilities

• Threat model and review LLM applications, RAG systems, inference pipelines, and agentic workflows.
• Identify prompt injection, insecure output handling, sensitive data exposure, excessive agency, and AI supply chain risks and drive mitigations.
• Assess security and privacy risks from AI vendors, foundation models, configurations, and ML libraries.
• Design red-team exercises for prompt injection, jailbreaks, model extraction, data poisoning, and tool or agent abuse.
• Document findings, reproduce failures, and collaborate with engineering teams on durable fixes.
• Define secure-by-default patterns for prompts, tools, retrieval, output handling, and model deployment.
• Establish trust boundaries, validation, and least-privilege access for agentic systems.
• Build guardrail libraries, security evaluation harnesses, model scanning, prompt scanning, and CI/CD security checks.
• Evaluate AI security tooling, including Garak, PyRIT, and Promptfoo.
• Support AI security incident detection, containment, triage, and root cause analysis.
• Contribute to policies and guidance, coach engineers, and translate threat intelligence into controls.

Required qualifications

• At least 7.5 years of experience in application security architecture and design.
• Experience in security engineering or application security with hands-on focus on AI and ML systems.
• Direct experience assessing LLM applications for prompt injection, insecure output handling, sensitive data exposure, and excessive agency.
• Proficiency in Python for security automation, evaluation scripts, and prototype tooling.
• Working knowledge of modern AI/ML stacks and at least one major cloud platform.
• Fifteen years of full-time education.

Preferred qualifications

• Familiarity with OWASP LLM Top 10, MITRE ATLAS, or NIST AI RMF.

Workplace

• The position is based at the Hyderabad office.

Source role information