Agentic AI Engineer in Springfield, Virginia at VAE, Inc.

VAE, Inc. is a full service IT Infrastructure Solutions Company focused on building, securing and supporting our clients’ mission critical enterprises. We provide a distinctive array of design, integration and implementation services as well as fully managed service offerings. VAE is at the forefront of leveraging multi-tenant capable technologies and shared IT services to create secure, reliable and cost-effective end-to-end services and solutions. We deliver exceptional infrastructure solutions with extremely talented employees using a client-focused partnering approach.

VAE, Inc. is seeking an Agentic AI Engineer to design, develop, and drive innovative agentic AI projects supporting national security programs and missions under the C5ISR program. This role will build reusable multi-agent orchestration patterns, define tool and service contracts, and partner directly with mission stakeholders to translate operational workflows into agent graphs and tool decompositions. The ideal candidate combines strong software engineering fundamentals with hands-on experience building and evaluating LLM-based agentic systems in mission-relevant environments.

VAE, Inc. is seeking an Agentic AI Engineer to design, develop, and drive innovative agentic AI projects supporting national security programs and missions under the C5ISR program. This role will build reusable multi-agent orchestration patterns, define tool and service contracts, and partner directly with mission stakeholders to translate operational workflows into agent graphs and tool decompositions. The ideal candidate combines strong software engineering fundamentals with hands-on experience building and evaluating LLM-based agentic systems in mission-relevant environments.

Key Responsibilities

• Design, develop, and drive innovative agentic AI projects supporting national security programs and missions.
• Build reusable multi-agent orchestration patterns using frameworks such as LangGraph, Strands, or AWS Bedrock AgentCore.
• Define and maintain the tool and service contracts agents call, including integration via emerging agentic protocols (MCP, A2A).
• Develop a shared evaluation harness for agent behavior, including trace analysis, regression testing, and quality metrics for non-deterministic outputs.
• Implement guardrails and control patterns: human-in-the-loop checkpoints, output validation, graceful failure handling, and cost and latency budgets.
• Partner with mission stakeholders and SMEs to translate operational workflows into agent graphs and tool decompositions.
• Perform prompt engineering, context design, and model selection against the authorized model inventory available in accredited environments.
• Contribute reusable components to a shared agent framework so that new mission agents can be stood up quickly and consistently.
• Interface with stakeholders to ensure deliverables meet organizational needs.
• Stay up to date with agentic AI research and emerging technologies relevant to mission needs.

• Bachelor's or Master's degree in Computer Science, Electrical Engineering, Data Science, Mathematics, or a related field.
• 4+ years of hands-on experience in software and/or AI systems development.
• Active Secret clearance, with the ability to obtain and maintain a Top Secret clearance.
• An active 8570 certification required
• Strong proficiency in Python.
• Hands-on experience building LLM-based applications, with at least one agent, RAG, or tool-calling system delivered.
• Working knowledge of at least one agent orchestration framework (LangGraph, Strands, Bedrock AgentCore, AutoGen, LlamaIndex, or comparable).
• Experience with API and tool integration, asynchronous workflow design, and state management.
• Ability to define, defend, and iterate on an evaluation approach for systems that do not produce deterministic output.
• Familiarity with cloud platforms (AWS preferred) and container technologies (Docker, Kubernetes).
• Strong collaboration and communication skills, including the ability to work directly with non-technical mission experts.

• Active Top Secret clearance.
• Experience working in national security or defense environments.
• Familiarity with classified environments and secure data handling practices.
• Familiarity with agentic interoperability protocols such as MCP, A2A, or AGNTCY.
• Experience with multimodal models applied to images, screenshots, or documents.
• Some exposure to cyber security work (DoD cyber operations, SOC, or security engineering), given the overall focus of the program.
• Experience with natural language processing (NLP), computer vision, or deep learning.
• Knowledge of tools like MLflow, Kubeflow, or Apache Airflow for model lifecycle management

VAE, Inc. is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to sex, race, color, religion, national origin, disability, protected Veteran status, age, or any other characteristic protected by law.