Description

Own observability, cost tracking, and security governance for AWS-based agent and MCP infrastructure.

Responsibilities

• Design and implement end-to-end observability using Dynatrace, integrating AWS CloudWatch, X-Ray, and Bedrock logs.
• Build granular cost models and automated reporting pipelines with anomaly detection for agent runtimes and vector databases.
• Develop distributed tracing and structured logging for LLM decision flows, tool usage, and sub-agent interactions.
• Evaluate and define target architectures for monitoring tools, including LangFuse, LiteLLM proxies, and AWS-native options.
• Design Attribute-Based Access Control (ABAC) architectures and scalable Terraform modules for identity and access management.

Required Skills

• 10+ years of experience in platform engineering or DevOps roles.
• Deep expertise with AWS services, specifically CloudWatch, X-Ray, and Bedrock.
• Hands-on experience implementing Dynatrace for observability and distributed tracing.
• Proficiency in Infrastructure as Code, particularly Terraform, for managing IAM and tagging strategies.
• Experience with Microsoft Teams and email integration for alerting and automated reporting.
• Strong understanding of distributed systems, including agent runtimes and MCP servers.
• Ability to evaluate and integrate third-party proxies like LiteLLM and LangFuse.
• Experience with Confluence for maintaining runbooks and documentation.

Preferred Skills

• Familiarity with the Cedar policy engine for fine-grained access control.
• Experience with ABAC design patterns and risk assessments.

Key Skills

Education

Any Gradute