Agentic AI Forward-Deployed Engineer

Booz Allen Hamilton•McLean, VA

7d•Remote

About The Position

As an AI-forward engineer, you know that intelligence has become a commodity. Every organization has access to the same frontier models and the advantage now belongs to those who can deploy them. Your ability to sit with a team, map how their work actually happens, and turn that understanding into reliable AI agents makes you an integral part of transforming how our programs deliver. We need your technical depth, advisory instincts, and ownership mindset to bring agentic AI to the programs where it matters most. You'll join our growing corps of forward deployed engineers within our Global Defense Disruption Cell, deploying to billable programs where people spend their days executing well-defined processes by hand. Your mission is to turn humans executing flow charts into agents executing flow charts, and to make those humans supervisors of that automation. You'll embed with program teams to map their workflows and exception paths, then exercise judgment about where deterministic software ends and model judgment begins. You'll build agents that are reliable, observable, and auditable from day one, and prove they work with evaluation suites before scaling them from shadow mode to production. Just as important, you'll bring the people along winning trust, managing the change, training the workforce, and leaving each program more capable than you found it. You won't do this alone. You'll operate with direct reach-back to the Disruption Cell, a team dedicated to clearing your blockers, escalating data-access and policy obstacles, and turning what you build in the field into reusable patterns, playbooks, and shared capabilities for the entire corps. The patterns you build once become capabilities every program can reuse. Work with us to define how AI transformation actually gets delivered one program, one workflow, one agent at a time. Join us. The world can’t wait.

Requirements

5+ years of experience in software engineering, machine learning engineering, technical advisory, or solutions engineering roles
2+ years of experience building applications with generative and agentic AI technologies, including LLMs, RAG, MCP, or multi-agent orchestration
Experience deploying an LLM-powered system or AI agent into a production environment, including evaluation, monitoring, and iteration after launch
Experience developing production code in Python or TypeScript
Experience integrating with enterprise systems and data sources, such as ERP, CRM, ITSM, or SharePoint platforms
Experience working directly with customers or business stakeholders to elicit requirements, map business processes, and guide adoption of new technology
Knowledge of evaluation techniques for non-deterministic systems, including golden datasets, regression testing, and human-in-the-loop feedback
Ability to operate independently in ambiguous, fast-moving environments and communicate with audiences ranging from engineers to executives
Ability to obtain a Secret clearance
Bachelor's degree

Nice To Haves

Experience with agentic coding and orchestration tools, such as Claude Code, Codex, Cursor, LangGraph, or the OpenAI Agents SDK
Experience in a forward deployed, embedded, or residency-style engineering role at a customer site
Experience with organizational change management, including training, communications, workforce transition, or user adoption programs
Experience with business process analysis or process mapping, including BPMN or value stream mapping
Experience with cloud platforms and containerized deployment, including AWS, Azure, Kubernetes, or Terraform
Ability to quantify automation impact in terms of cost savings, risk mitigation, or revenue uplift
Secret clearance
Master's degree

Responsibilities

Embed with program teams to observe and document how work actually happens, including workflows, systems, data flows, decision points, and the exception paths that never make it into the documented process.
Produce operating maps and automation roadmaps that prioritize high-volume workflows by expected value and risk, and decide where deterministic software, agent judgment, and human-in-the-loop approval each belong.
Design, build, and deploy AI agents that reason, plan, and act across the program's existing tools, APIs, and data sources. Build on the systems teams already use rather than forcing a migration.
Develop evaluation frameworks, golden datasets, and regression tests that turn non-deterministic behavior into evidence, and scale agents from shadow mode to increasing autonomy to production.
Instrument everything including audit trails, monitoring, and metrics that let program leadership see exactly what agents are doing and what value they deliver in hours saved, risk reduced, and outcomes improved.
Drive adoption and change management, train the workforce, redesign roles around supervision of automation, and de-risk the transformation for the people living through it.
Codify what you learn into playbooks, reusable components, and field feedback for the Disruption Cell, and escalate blockers through direct reach-back so no engagement stalls on access or politics.