Glossary: AI, agents and JD Edwards, explained clearly

An open protocol that lets a language model use external tools and data sources in a controlled way. An MCP server defines what the AI can query and with what limits; it is the piece that makes it possible to ask JD Edwards questions in natural language without exposing the database.

An EnterpriseOne component (Tools 9.2) for building orchestrations: processes that read and write in JDE combining forms, rules and services. It is Oracle’s recommended path for automation and for connecting the ERP with AI services.

JD Edwards’ REST services layer exposing forms and orchestrations to external applications while respecting ERP security. The technical bridge between AI agents and EnterpriseOne.

Software that uses a language model to interpret a request, decide steps and execute actions through tools (like orchestrations or MCP queries), with defined permissions and auditing.

A technique combining LLM generation with retrieval of your company’s own documents, so answers are grounded in your policies, contracts or manuals — not just the model’s general knowledge.

Oracle’s cloud. It offers generative AI, vision and document-understanding services that Orchestrator can invoke from JDE, plus infrastructure to host EnterpriseOne.

Mexico’s mandatory electronic invoice. Its issuance, reception and validation can be automated and integrated with the ERP.

An ERP whose access and automation flow through AI agents: users query and execute in natural language while the system of record and its rules remain intact.