This topic covers enterprise AI agent platforms and agent frameworks used to automate, observe and orchestrate supply‑chain and operations workflows. By 2026, organizations are combining large language models, retrieval‑augmented generation (RAG), multimodal perception and orchestration layers to handle tasks such as inventory audits, exception triage, cross‑system order routing and predictive maintenance. Key categories include AI automation platforms (cloud‑managed model and orchestration stacks) and agent frameworks (developer and no‑/low‑code runtimes for building autonomous workflows). Representative tools illustrate the stack: platform services such as Vertex AI and Cohere provide managed LLMs, embeddings and model hosting; RAG and document agent tooling like LlamaIndex turn unstructured operational data into queryable knowledge; orchestrators such as IBM watsonx Orchestrate and Oracle’s agent offerings coordinate actions across ERP/WMS/TMS systems; stack and no‑code vendors like StackAI and Anakin.ai speed agent composition and governance for non‑developers; specialized offerings—Yellow.ai for CX/EX agents and Gather AI for intralogistics—surface domain capabilities (multichannel agents and autonomous drone+computer‑vision audits, respectively). Current trends reflected in vendor functionality include tighter integration between edge sensor feeds and LLM decision layers, stronger emphasis on traceability and guardrails for agent actions, and hybrid deployment models to meet latency and privacy requirements. Adoption drivers are operational cost pressure, the need for 24/7 exception handling, and maturing standards for agent governance. Evaluations should prioritize integration with existing supply‑chain systems, support for RAG and multimodal inputs, observability, and enterprise controls for security, rollback and human‑in‑the‑loop interventions.