EPAM Systems
Website:
epam.com
Job details:
We are seeking a Senior System Engineer with deep expertise in AWS IAM, Python, and Agentic AI to architect and secure next-generation AI systems. In this role, you will design secure-by-design controls for agentic mesh AI environments, govern multi-account AWS landscapes, and enable policy-aware AI agents operating within defined security and compliance boundaries.
Responsibilities
- Design and implement secure-by-design controls for agentic mesh AI systems built on Amazon Bedrock
- Define trust boundaries between autonomous AI agents and services
- Assess and mitigate risks related to prompt injection, model misuse, and agent-to-agent trust and verification
- Address data leakage and cross-agent contamination concerns across AI workloads
- Enable policy-aware AI agents that operate within defined security and compliance constraints
- Apply network security patterns for AI workloads, including VPC design, private endpoints, and Bedrock access controls
- Architect and govern multi-account AWS environments leveraging Organizations and Control Tower
- Implement account isolation strategies for AI workloads, data, and agent execution
- Enforce security controls using AWS Security Hub, GuardDuty, Config, and CloudTrail
- Collaborate across AI engineers, cloud teams, and governance stakeholders to align on security architecture
- Produce strong architectural documentation and communicate complex security concepts clearly
Requirements
- 5-8 years of general IT experience, including 5+ years of professional experience in system engineering and cloud security
- Expertise in AWS IAM, Amazon AWS Security, and AWS Security Hub
- Proficiency in Python for automation and security tooling
- Background in Amazon Bedrock AgentCore and AI Agents Frameworks
- Knowledge of AI & ML Strategy with focus on secure deployment patterns
- Skills in multi-account AWS governance using Organizations and Control Tower
- Familiarity with AWS security services, including GuardDuty, Config, and CloudTrail
- Understanding of network security patterns for AI workloads, including VPC design and private endpoints
- Capability to act as a structured thinker with a security-first mindset and strong communication skills
- Proficient communication skills in English (B2 level or higher)
Nice to have
- Understanding of agentic AI concepts and design principles
- Familiarity with multi-agent systems architecture
- Knowledge of tool-using and autonomous agents
Click on Apply to know more.