AI Agent Engineer

Client: Client•Reston, VA

About The Position

The AI Agent Engineer will lead the implementation, administration, functional and security validation, and operational management of enterprise coding-agent capabilities. This senior/lead-level role requires hands-on experience operating AI coding technologies in enterprise production environments, as well as the ability to work independently, troubleshoot complex integrations, establish technical standards, and provide technical leadership to engineering teams adopting coding agents. The role focuses on enterprise coding-agent platforms and workflows, agent engineering and orchestration, secure Model Context Protocol (MCP) integrations, identity and access management, AWS AI platform capabilities, and Python development. The engineer will assess agent behavior from functional and platform-administration perspectives and validate that agents, tools, identities, and credentials operate within appropriate security boundaries. The position also involves developing and validating agent services, integrations, automation utilities, APIs, and testing frameworks. The preferred candidate has led complex implementations and has production experience administering developer-facing platforms used by multiple engineering teams.

Requirements

Deep hands-on experience with OpenAI Codex or coding agents and GitHub Copilot, including implementation, configuration, administration, troubleshooting, and enterprise deployment.
Experience configuring and administering enterprise policies, user and group access, repository access, agent permissions, model configuration, tool access, policy enforcement, usage controls, and security controls.
Experience implementing and supporting coding-agent workflows across the software development lifecycle.
Hands-on experience designing, configuring, and building AI agents, including agent orchestration, custom agents or assistants, tools and integrations, prompt and instruction architecture, context management, and tool execution.
Strong working knowledge of Model Context Protocol (MCP), including experience building and securing MCP servers.
Understanding of MCP scopes and permissions, authentication and authorization, fine-grained authorization, identity propagation, credential handling, and secure access to enterprise APIs and systems.
Deep understanding of enterprise identity and access management, with hands-on experience implementing several of SSO, SAML, OAuth 2.0, OIDC, identity providers, SCIM, AWS IAM, AWS IAM Identity Center, Active Directory or Entra ID, bearer tokens, JWTs, service identities, RBAC, and fine-grained authorization.
Ability to distinguish authentication from authorization and design identity flows for human users and autonomous agents.
Strong working knowledge of AWS, including hands-on experience with AWS services and infrastructure such as IAM roles and policies, ECS, ALB, Lambda, API Gateway, CloudWatch, Secrets Manager or Parameter Store, and VPC/networking fundamentals.
Strong proficiency in Python.
Ability to develop REST APIs, agent services, automation utilities, integration services, test harnesses, and validation frameworks.
Senior/lead-level ability to work independently, troubleshoot complex integrations, establish technical standards, and provide technical leadership.
Production experience operating AI coding technologies in enterprise environments is expected; experience solely developing proofs of concept does not match the stated profile.

Nice To Haves

Experience with Amazon Bedrock, Amazon Bedrock Agents, and Amazon Bedrock Agent Core.
Experience with asynchronous Python, API development frameworks, and structured logging.
Experience designing security boundaries around MCP tools.
Experience implementing least-privilege access for AI agents.
Experience developing automated agent evaluation frameworks, test harnesses, or regression suites.
Shell scripting or Bash experience.
PowerShell experience.
Prompt-injection and AI security testing experience.
Significant recent hands-on experience with GitHub Copilot or comparable enterprise AI coding platforms.
Hands-on experience with OpenAI Codex or production coding-agent technologies.
Multiple years of experience engineering and operating enterprise developer platforms, cloud platforms, or AI/ML platforms.
Production experience administering developer-facing technology used by multiple engineering teams.
Evidence of leading technically complex implementations rather than participating solely as an individual application developer.

Responsibilities

Lead implementation, administration, functional validation, security validation, and operational management of enterprise coding-agent capabilities.
Implement, configure, administer, and troubleshoot AI coding agents and developer AI platforms, including enterprise deployments.
Configure platform policies, user and group access, repository access, agent permissions, model and tool access, usage controls, and enterprise security controls.
Implement and support coding-agent workflows across the software development lifecycle.
Design, configure, and build AI agents, custom assistants, agent tools, integrations, and orchestrated workflows.
Build and secure MCP servers and establish appropriate scopes, permissions, identity propagation, credential handling, and security boundaries for MCP tools.
Design identity and access flows for human users and autonomous agents, applying least-privilege access principles.
Develop REST APIs, agent services, automation utilities, integration services, test harnesses, and validation frameworks using Python.
Validate security and authorization controls, including repository access, tool permissions, token handling, privilege escalation, secrets exposure, cross-user or cross-project access, and prompt-injection or tool-abuse scenarios.
Troubleshoot complex integrations, establish technical standards, and provide technical leadership to engineering teams adopting coding agents.
Develop or support automated agent evaluation frameworks, test harnesses, and regression suites.