AI Systems Engineer

KBR•Chantilly, VA

3d•Onsite

About The Position

KBR is seeking a visionary AI Systems Engineer to join our team, providing critical technical services to the U.S. Government (USG). This role supports a dedicated USG Program Office, in shaping, designing, and deploying cutting-edge AI and Machine Learning (AI/ML) initiatives for national security space architectures. As an AI Systems Engineer, you will serve as both a hands-on technical architect and a strategic advisor. You will leverage the Government customer’s massive enterprise datasets, spanning complex program budgets, scheduling, and financials, to build proof-of-concept models (including Large Language Models and financial predictive tools). You will lead the evolution of the AI roadmap by identifying cutting-edge commercial technologies and directing successful prototypes into robust, production-ready enterprise tools.

Requirements

Active TS/SCI with Polygraph.
Bachelor’s degree with 5+ years of relevant experience, Master’s degree with 3+ years of relevant experience, or PhD with 0+ years of relevant experience.
Strong foundations in AI/ML concepts, natural language processing (NLP), and statistical modeling.
Demonstrated experience building prototypes using Python, R, or similar tools, specifically with frameworks related to LLMs (e.g., llama.cpp, LangChain, Hugging Face) or data analytics.
Experience working with complex enterprise datasets (e.g., financial, resource planning, schedule networks, or programmatic data).
Exceptional ability to translate highly technical AI concepts—such as LLM training methodologies, RAG architecture, model alignment, and corpus curation—into strategic, actionable business recommendations for non-technical executives and government leaders. This includes clearly articulating the compute costs, security risks, data provenance, and ongoing maintenance requirements involved in building versus adopting language models.

Nice To Haves

Experience applying ML or advanced statistical modeling to predictive financial analysis, budget execution, risk modeling, or project scheduling.
Understanding of the challenges of deploying AI/ML models in high-security, air-gapped government cloud environments (e.g., AWS C2S).
Prior experience in technology scouting, venture capital technical vetting, or acting as an industry-to-government technology transition agent.
Familiarity with the DoD/IC acquisition lifecycle, systems engineering processes, and SETA advisory roles.

Responsibilities

Help shape and define the AI/ML roadmap for the Government customer office.
Guide the government on how to securely build, adopt, govern, and maintain foundation models and custom LLMs across sensitive environments, ensuring robust model lifecycle management.
Design and build hands-on, high-impact proof-of-concept models. This includes training, fine-tuning, and leveraging Large Language Models (LLMs) for search, summarization, and business intelligence, as well as developing advanced financial, budget, and scheduling analysis models.
Work directly with massive, unstructured and structured enterprise datasets (financials, acquisition timelines, schedules, program execution metrics). Clean, curate, and pipeline this data to create high-quality training corpora, instruction-tuning datasets, and knowledge bases for Retrieval-Augmented Generation (RAG) to discover trends, anomalies, and optimization opportunities.
Actively scout the commercial tech sector, academia, and open-source communities for breakthroughs in generative AI, NLP, and predictive analytics. Curate a pipeline of high-performing external groups and technologies for USG consideration.
Formulate technical recommendations, draft statements of work, and establish the system integration requirements needed to bring external contractors on board to scale and operationalize your prototypes.
Provide objective technical evaluation of contractor proposals, designs, and deliverables to ensure they align with government AI safety, security, and performance standards. This includes overseeing LLM red-teaming, bias testing, hallucination mitigation, and ongoing alignment evaluations.