Enterprise agent orchestration platforms coordinate retrieval-augmented generation (RAG), tool use, memory and multi-step agents to automate complex business workflows. This topic covers the vendor platforms, open frameworks and marketplaces that organizations use to design, test, deploy and govern autonomous agents at scale. Key trends driving adoption include the move from isolated LLM experiments to stateful, observable agent pipelines; demand for enterprise controls (access, auditing, data lineage); and options for managed vs. self-hosted execution. Platform types fall into complementary categories: agent frameworks and engineering libraries (e.g., LangChain’s engineering platform and LangGraph for stateful orchestration) provide primitives for building and evaluating agentic applications; cloud ML suites (e.g., Google’s Vertex AI) supply model discovery, training/fine-tuning, and production deployment; low-code/no-code builders (MindStudio, Lindy, Anakin.ai) let business teams compose agents and automation visually; and developer-focused runtimes and marketplaces (GPTConsole, AutoGPT) focus on SDKs, CLIs, event chaining and lifecycle management. Emerging marketplaces and tool catalogs help teams discover, share and monetize prebuilt agents and integrations. As of 2026-03-23, enterprises prioritize platforms that integrate RAG with secure data access, observability, and governance while enabling hybrid deployment models. Differences between offerings center on developer ergonomics, state management, lifecycle tools, and the balance between prebuilt apps and engineering flexibility. Evaluations should weigh integration with existing data systems, controls for sensitive data, and the operational features—testing, evaluation, and rollback—that determine whether agentic workflows can be reliably operated at enterprise scale.