Full Stack Engineer

Build Ai Systems That Drive Real-World DecisionsMost AI projects never make it past a demo.Most enterprise software never makes it past a dashboard.This opportunity is different.We're building a next-generation logistics, maintenance planning, and decision-support platform that helps maintenance planners anticipate risk, optimize schedules, manage inventory, and improve fleet readiness before problems impact operations.This is not a greenfield project.The foundation already exists. Core inventory systems, maintenance planning workflows, forecasting visualizations, reporting capabilities, and compliance requirements have already been established. We are entering the phase where advanced AI, machine learning, forecasting, optimization, and agentic workflows begin creating real operational value.What You'll DoDevelop workload forecasting, schedule-risk, optimization, and simulation models using Python, NumPy, SciKit-Learn, PyTorch, TensorFlow, or appropriate statistical methods.Build production features that connect model outputs to backend APIs, databases, and frontend visualizations.Implement Monte Carlo simulations, schedule optimization heuristics, workload leveling, uncertainty modeling, and critical-path analysis.Integrate multimodal embeddings, semantic search, vector databases, and retrieval-augmented workflows.Build agentic workflows using Pydantic AI, CopilotKit, AG-UI, MCP integrations, and tool-calling patterns.Collaborate with full-stack engineers to expose model results through FastAPI, PostgreSQL/pgvector, and React/TypeScript dashboards.Develop user-facing planning, forecasting, and decision-support features for maintenance scheduling, parts availability, AWP constraints, and workforce planning.Write tests and validation workflows for model behavior, backend APIs, and user-facing analytics.Support deployment of AI and forecasting features into containerized, Kubernetes-based, and potentially air-gapped environments.What You'll Be BuildingYou won't be creating models that sit in notebooks.You'll be helping build production systems that combine:Forecasting and predictive analyticsSchedule optimization and simulationInventory intelligenceSemantic search and retrievalAgentic AI workflowsReal-time operational dataUser-facing decision support toolsThe platform already includes inventory management, maintenance planning, forecasting visualizations, reporting, and modern full-stack infrastructure.The next phase focuses on adding intelligence:AI-guided recommendationsSemantic retrieval and vector searchOptimization enginesAgent tool-calling workflowsReal-time data pipelinesAdvanced forecasting and simulation capabilitiesTechnology StackFrontendReactTypeScriptModern planning and visualization interfacesInteractive forecasting and scheduling viewsBackendPythonFastAPIPostgreSQLSQLAlchemypgvectorAI / Machine LearningPyTorchTensorFlowSciKit-LearnEmbeddingsSemantic SearchRetrieval-Augmented Generation (RAG)Pydantic AICopilotKitAG-UIAgent Tool CallingInfrastructureDockerKubernetesContainerized DeploymentsSecure Government EnvironmentsAir-Gapped SystemsRequired QualificationsB.S. degree in Computer Science, Engineering, Mathematics, Statistics, Operations Research, Data Science, or a related technical field.U.S. Citizenship and active Secret Clearance, or ability to obtain one where applicable.Strong proficiency in Python 3, NumPy, Pandas, and applied machine learning or statistical modeling.Experience with at least one major ML framework such as PyTorch, TensorFlow, or SciKit-Learn.Experience building production software beyond notebooks, including APIs, databases, tests, and deployment workflows.Familiarity with FastAPI, Pydantic, PostgreSQL, and modern backend development practices.Ability to translate domain constraints into practical forecasting, simulation, or optimization logic.Preferred QualificationsExperience with schedule optimization, operations research, Monte Carlo simulation, discrete-event simulation, or resource-constrained planning.Experience with PostgreSQL, pgvector, embeddings, semantic search, and retrieval-augmented generation.Familiarity with Pydantic AI, CopilotKit, AG-UI, MCP integrations, agentic coding, and agent harness utilization.Experience with React, TypeScript, TanStack Query, data visualization, dashboards, or full-stack feature development.Experience with Kubernetes, Helm-style deployments, Docker, and air-gapped environments.Knowledge of Navy maintenance processes, submarine systems, shipyard operations, depot maintenance, logistics, parts management, or workforce scheduling.Experience with model evaluation, experiment tracking, reproducibility, observability, and production monitoring.Why This Role Is DifferentMany machine learning roles stop at prediction.Many software roles stop at visualization.This role is about helping people make better decisions.You'll work on forecasting, optimization, inventory intelligence, semantic retrieval, and AI-assisted workflows that transform complex operational data into actionable recommendations.The foundation is already built.The challenge now is turning data into insight, insight into decisions, and decisions into operational outcomes.