Agentic AI Prototyping Engineer

Aimlroles; North Dakota; Hybrid; USD 99,000 - 225,000; Full time

Job summary

Booz Allen is seeking an experienced Agentic AI Engineer to define, design, and deliver mission-aligned agentic AI prototypes in rapid, four-week cycles. You will work with product leads, architects, and engineers to build, test, and deploy agent-based solutions that balance performance, reliability, and cost.

Responsibilities include designing agent loops, memory management, and retrieval workflows, while applying best-fit frameworks and ensuring end-to-end delivery from concept through

Qualifications

• Experience building agentic AI applications in production or prototype environments.
• Experience implementing agent orchestration frameworks, including AWS Strands.
• Experience designing and implementing agent loops, tool-calling workflows, and multi-agent interaction patterns.
• Knowledge of LLM platforms and providers, including AWS Bedrock, OpenAI, or Ollama.

Responsibilities

• Define, design, and deliver agentic AI prototypes in fast, iterative cycles.
• Collaborate with product leads, solution architects, and engineers to build and deploy agent-based solutions.
• Apply best-fit frameworks and patterns for agent orchestration, memory, and retrieval.
• Balance solution quality, reliability, and cost in development from concept to delivery.

Skills

Agentic AI

Agent orchestration

LLM platforms

Prompt engineering

External tool integration

Memory management

End-to-end delivery

Communication

Education

HS diploma or GED

CS/Engineering degree

Tools

AWS Strands

AWS Bedrock

OpenAI

Ollama

Tool-calling workflows

Job description

Booz Allen is seeking an experienced Agentic AI Engineer to define, design, and deliver mission-aligned agentic AI prototypes in rapid, four-week cycles. You will work with product leads, architects, and engineers to build, test, and deploy agent-based solutions that balance performance, reliability, and cost.

Responsibilities include designing agent loops, memory management, and retrieval workflows, while applying best-fit frameworks and ensuring end-to-end delivery from concept through