Principal AI Architect

Analytica•Bethesda, MD

About The Position

We are seeking an experienced Principal AI Architect to set the technical direction for safe, traceable, human-in-the-loop agentic AI and analytics for a Federal data analytics and AI modernization initiative. The team will deliver a secure, scalable platform that integrates structured and unstructured data, provides data visualization and traceable AI-assisted analytics, and gives examiners centralized tools for search, review, monitoring, and decision support. The platform will support professional judgment and will not replace authoritative agency financial or grants-management systems or execute financial transactions.

Requirements

Deep experience designing agentic AI, LLM, NLP, semantic-search, or advanced analytics solutions.
Demonstrated experience defining AI architecture, model strategy, retrieval patterns, evaluation methods, guardrails, and governance.
Experience leading Federal delivery of secure AI capabilities using Government-controlled data and infrastructure.
Strong knowledge of RAG, grounded generation, source traceability, human-in-the-loop design, and on-premises model deployment.
Strong communication, documentation, problem-solving, and cross-functional collaboration skills.
Familiarity with relevant project technologies such as Linux, PostgreSQL or Microsoft SQL Server, Java or .NET, Jenkins, self-hosted Azure DevOps, Kubernetes/Rancher, and open-source AI models.
U.S. citizenship and ability to obtain and maintain Top Secret eligibility, as required for all contractor personnel supporting the effort.
Active Top Secret highly preferred.

Nice To Haves

Experience supporting Federal financial management, budget execution, grants management, payment systems, award oversight, financial reporting, or financial stewardship is preferred.
Experience supporting Federal financial management, grants, payment, or award-oversight platforms.

Responsibilities

Perform role-specific work to set the technical direction for safe, traceable, human-in-the-loop agentic AI and analytics.
Collaborate with pilot agencies, and multidisciplinary technical teams through iterative development and user testing.
Produce clear documentation that supports Government operation, maintenance, transition, and future expansion.
Follow Federal security, privacy, accessibility, records-management, and Government data-ownership requirements.