Principal Developer, Agentic AI

Overview ABOUT JAGGAER JAGGAER provides an intelligent Source-to-Pay and Supplier Collaboration Platform that empowers organizations to manage and automate complex processes while enabling a highly resilient, responsible, and integrated supplier base. With 30 years of expertise, we specialize in solving complex procurement and supply chain challenges across various industries. Our 1,300+ global employees are obsessed with ensuring customers get full value from our products - ultimately enhancing and transforming their businesses. For more information, visit www.jaggaer.com JAGGAER is an AI-powered procure to pay and source to pay software solutions platform for manufacturing, public sector and higher education. We are seeking a hands-on Principal Engineer to lead the architecture and evolution of our enterprise agentic AI platform. You will work across AI agents, multi-agent orchestration, workflows, tools, enterprise integrations, retrieval, model providers, micro-frontends, observability, security, and governance. Your responsibility will be to create a coherent, reusable platform that enables multiple product teams to build and operate secure, reliable AI capabilities at scale. This is not a documentation-only architecture role. You will be expected to design systems, review and write production code, resolve complex technical issues, guide platform migrations, and mentor senior engineers. Principal Responsibilities Define the target architecture for enterprise AI agents, workflows, orchestration runtimes, tools, retrieval, model integrations, and embedded user experiences. Establish clear platform contracts for agent definitions, versions, advertised capabilities, tools, workflows, sessions, memory, and runtime execution. Design and evolve multi-agent and Supervisor Agent architectures that can safely coordinate worker agents, child supervisors, tools, and long-running workflows. Define shared runtime operations such as invoke, stream, resume, approve, cancel, status, events, and trace. Ensure secure propagation of tenant, user, authorization, run, and delegation context across distributed agent and workflow calls. Build deterministic security, policy, approval, budget, retry, and lifecycle controls around probabilistic model decisions. Guide micro-frontend and embedded AI architecture, including streaming, orchestration progress, tool activity, evidence, approvals, and error handling. Define provider-independent model integrations and model-selection strategies based on quality, latency, reliability, and cost. Establish agent evaluation systems covering capability selection, plan quality, grounding, hallucination, safety, latency, cost, and user experience. Drive observability across model calls, agents, tools, workflows, retrieval, evaluation, and final synthesis. Design for retries, idempotency, partial failures, cancellation, long-running execution, human pauses, and process recovery. Lead…