Additional Requirements
Must meet eligibility requirements for U.S. Government Security Clearance (they do not need to currently hold one)
Ability to travel onsite to Norfolk, VA (estimated 1-2 times over the project duration)
- Location: Remote US | Ability to travel onsite to Norfolk, VA (estimated 1-2 times over the entire project duration)
PLEASE FOCUS ON CANDIDATES WITH GCP EXPERIENCE (Google Cloud architecture certification (e.g., Professional Cloud Architect) is a huge plus)
Project Description
The customer is a manufacturer supporting U.S. Navy submarine production, pursuing a strategic modernization of its manual, paper-based shop-floor operations. The business goal is to replace fragmented, paper-driven manufacturing processes with a proprietary AI-enabled workforce readiness platform that gives Foremen, Technicians, QC Inspectors, and Program stakeholders AI-generated work instructions, smart technician assignment, and real-time quality and compliance visibility. The engagement is delivered in two phases across six workstreams: a FedRAMP High/IL4-compliant Google Cloud Platform Landing Zone that establishes the secure cloud foundation, followed by platform development spanning a manufacturing data integration engine, data connector migration, Looker analytics, and a workforce enablement platform powered by four AI agents. Delivery is distributed across multiple delivery locations to balance the teams and specialized skill availability, all within the strict US-persons-only access controls required once the customer's Controlled Unclassified Information (CUI) is in use.
Main Responsibilities
Platform Architecture & Technical Design: owns the platform's architecture and Technical Design Document
Owns the platform architecture and Technical Design Document, spanning the platform's role-based UI, RAG pipeline, and four AI agents.
Leads an early Assess & Design phase that all downstream platform build and AI/agent development work depends on.
API Design & RBAC/IAM Model: designs the platform's API layer and access-control model
Designs the platform's API layer and its role-based access control (RBAC) / IAM model, mapped explicitly to FedRAMP High/IL4 requirements.
Coordinates the RBAC/IAM design with the Network/Security Specialist to ensure it holds up through the CUI cutover and beyond.
KPI Tracking Framework: owns the framework that measures program success
Owns implementation of the KPI Tracking Framework defined during Discovery, turning the program's agreed success measures into a built, measurable capability during a subsequent build phase.
Ensures the framework can track improvement in workforce decision-making, supply chain risk reduction, and machinery reliability, the program's three original presales KPI targets.
Cross-Workstream Technical Authority: is the technical authority the platform Development Team builds against
Remains engaged across the full life of the program as the technical authority guiding the Full Stack Developer, AI/ML Engineer, Front End Developer, Back End Developer, and Mobile Developer.
Ensures the platform is architected with the customer's long-term expectation in mind that it evolves into a reusable product across the broader defense shipbuilding sector, not a single-client build.
Skills
Must-have
5 8+ years of solution/platform architecture experience, including AI-enabled enterprise platforms
API design experience for multi-persona, role-based platforms
RBAC/IAM design experience, including compliance mapping to frameworks like FedRAMP High/IL4
Experience defining or implementing KPI/measurement frameworks tied to business outcomes
Demonstrated ability to serve as end-to-end technical authority across a multi-role, multi-agent platform build
US-person status, consistent with the CUI/FedRAMP High/IL4 access restrictions governing this engagement
Nice-to-have
Experience architecting RAG-based or agentic AI systems
Google Cloud architecture certification (e.g., Professional Cloud Architect)
Manufacturing, industrial, or DoD-adjacent domain experience
Experience architecting platforms intended for reuse across multiple clients