You are:

A Context Engineer, pivotal in orchestrating the delivery of AI-driven solutions - bridging business analysis, solution design, and integration to ensure AI has the right context to produce consistently strong, mission-aligned results. You translate approved user stories into AFS-compliant agent prompts, designs and governs the subagents that execute them, operates the AI build pipeline, and owns a high-performing prompt library. You integrate seamlessly with CI/CD pipelines while upholding security and ATO (Authority to Operate) compliance standards.

The work:

• Translate approved user stories into AFS-compliant agent prompts, ensuring alignment with project objectives and mission outcomes.
• Design the architecture for specialized subagents that operate within the build pipeline - defining each agent's purpose, scope, tools, and boundaries.
• Build routing, delegation, and escalation logic that determines when the system answers directly, invokes a subagent, runs specialists in parallel, asks a clarifying question, uses a tool, escalates to a human, or declines an unsafe action.
• Define agent-to-subagent contracts: inputs, context boundaries, structured outputs, confidence and evidence, latency budgets, cancellation behavior, and failure handling.
• Configure, manage, and operate AI build pipelines using generative agentic tools and orchestration frameworks (e.g., Claude, Codex, LangGraph).
• Own and maintain the prompt library including version control, quality scoring, and continuous improvement - refining prompts based on reviewed outputs and feedback.
• Create and maintain RAG knowledge bases and query prior context files for precedent and predictive insight before authoring new prompts.
• Oversee and integrate CI/CD pipelines, embedding security validations (SAST, vulnerability scanning) and automated ATO evidence generation for seamless compliance.
• Build rapid prototypes using AI-assisted coding to test complex prompting and data workflows before full implementation.
• Document the work and enable others - producing clear write-ups and becoming the team's expert and trainer on prompting and context practices.
• Collaborate with cross-functional teams including AI Product Leads, Senior AI Engineers, and Verification Engineers to deliver AI solutions successfully.

Here’s what you need:

• Proven experience in business analysis, solution integration, or solution design.
• Strong background managing and configuring AI build pipelines with generative, agentic AI tooling and agent-orchestration frameworks - such as Claude, Codex, or LangGraph.
• Mastery of LLM prompting with hands-on API experience - command of prompt parameters, components, and conversational LLM outputs.
• Familiarity with CI/CD processes and tools, including security validation and compliance integration.
• Ability to query and analyze structured data and prior context files to derive actionable insights.
• Understanding of AI build pipelines, subagent orchestration, and RAG knowledge-base creation.
• Familiarity with version control systems and managing a prompt library.
• Experience working in agile environments or with scrum methodologies.
• Knowledge of security and compliance processes within the context of AI operations.

Bonus points if you have:

• Bachelor's degree in Business, Information Systems, Computer Science, Engineering, or equivalent practical experience.
• Basic coding skills for rapid prototyping and iterating on data/prompting workflows.
• Multimodal prompting experience (images, video) and familiarity with text-embedding models.
• Copywriting, data science, or software engineering background.
• Prior work with conversational LLM outputs and feedback-driven prompt refinement.
• Excellent written communication - able to document work clearly and train others.

Eligibility Requirements:

• U.S. Citizenship required.
• Ability to work with federal clients and comply with applicable AFS, client, security, and program requirements.
• Ability to work in a hybrid, remote, or client-site environment as program needs require.