Senior AI Engineer (Agentic Systems)

Company Overview Pantheon Data (a Kenific Holding company) is a private, small business based in the Washington, DC, area. Pantheon Data was founded in 2011, initially providing acquisition and supply chain management services to the US Coast Guard. Our service offerings have grown in the past ten years, including infrastructure resiliency, contact center operations, information technology, software engineering, program management, strategic communications, engineering, and cybersecurity. We have also grown our customer base to include commercial clients. The company has used this experience to expand our service offerings to other agencies within the Department of Homeland Security (DHS), the Department of Defense (DoD), and other Federal Civilian Agencies. Position Overview Pantheon Data is seeking a Senior AI Engineer (Agentic Systems) to design, build, and operate the agentic framework at the core of our AI platform in AWS GovCloud, serving workloads that process CUI, PII, and PHI. You will own the framework that routes requests to the right agents, composes new agents at runtime from a registry of capabilities, and coordinates their tools, state, and handoffs, so that these systems behave predictably, securely, and auditably under FedRAMP High and DoD IL4/IL5. This role owns the agent runtime: orchestration, routing, the capability registry, and the contracts that capabilities plug into. Capabilities such as retrieval, document extraction, and application integrations are built by machine learning and software engineers and exposed through the registry; you define how they are described, discovered, composed, and governed. You will partner closely with the Data Engineer, who owns the pipelines and data stores agents draw on; the Cloud Security Engineer, who owns cloud security controls and ATO evidence; and the DevSecOps and Cloud Operations Engineers, who own delivery and infrastructure. This is a systems engineering role, not a prompt-only or prototype role: successful candidates can walk through production agentic systems they have personally built, including the orchestration design, how routing decisions were made and evaluated, how failures were contained, and what they would change. Responsibilities Agent Orchestration: Design and build the orchestration layer that plans, dispatches, and coordinates work across agents, including multi-step and multi-agent workflows, handoffs, parallel execution, cancellation, and human-in-the-loop checkpoints. Routing: Build the routing that maps requests and intermediate steps to the right agents, tools, and models, combining deterministic rules, classifiers, and LLM-based routing where each is appropriate; measure routing accuracy and cost, and make routing decisions observable and explainable. Capability Registry and Dynamic Agents: Own the capability registry, including schemas for describing capabilities, versioning, discovery, and access policy, and the mechanisms that assemble agents at runtime from registered capabilities, tools, and instructions. Tool and Integration Contracts: Define the interfaces capabilities plug into, including tool definitions, Model Context Protocol (MCP) servers, input and output schemas, and error semantics, and help capability owners onboard to them. State, Memory, and Reliability: Design agent state, session memory, and context management; build retries, timeouts, idempotency, loop and budget limits, and graceful degradation so agent failures are contained and recoverable. Evaluation and Observability: Build evaluation harnesses for agent behavior, including routing accuracy, task completion, tool-call correctness, and regression across model and prompt versions, and trace every agent run end to end for debugging, cost, and latency. Agent Security and Auditability: Enforce least-privilege tool access per agent and per request, contain prompt injection across tool and agent boundaries, and record a complete audit trail of agent decisions and acti…

View the full posting and apply