JD Edwards: An Agentic Enterprise

The Digital Workforce does not write directly to the JD Edwards database. It uses JD Edwards AIS to call the relevant application functionality, meaning existing JD Edwards application logic and security controls continue to apply. 

For example, in the Work Order Time Entry demonstration, the assistant validated the work order and required information before performing the update in JD Edwards. This allows users to simplify the interaction without bypassing the controls already built into the ERP. 

Yes. Email-driven workflows can be used to trigger assistants and agents based on inbound requests, such as processing documents or responding to operational requests. 

The platform can also be accessed through other conversational interfaces, including Teams, Copilot and compatible AI interfaces, providing flexibility in how users and systems initiate requests. 

For autonomous processes, organisations can define validation requirements and confidence thresholds before an agent is permitted to perform an action in JD Edwards. 

Licensing depends on the size and requirements of the organisation, so there is not a single standard price that applies to every customer. 

Fusion5 typically begins with a proof-of-concept or proof-of-value phase before moving into production and determining the appropriate licensing model. Specific pricing is therefore discussed based on the customer's environment, intended use cases and scale of deployment. 

Initial onboarding can be completed quickly. 

The Digital Workforce is delivered as a SaaS platform and requires connectivity to the customer's JD Edwards environment, along with configuration for SSO, security and trusted connections. 

Depending on the customer's environment and how quickly the required teams can work together, the initial platform connection can typically be completed within a few days to less than a week. 

From there, organisations can begin with existing experts and templates before developing customer-specific assistants and agents. 

Yes. The platform is designed to understand customer-specific JD Edwards environments rather than assuming every implementation is standard. 

Fusion5 described the ability to reverse-engineer custom objects, category codes, UDCs and other customer-specific configurations and capture that information within a semantic layer or ontology. 

That organisational context is then used to improve the accuracy and relevance of responses. 

Yes. At the time of the webinar, approximately seven standard languages were available within the platform. 

Individual users can configure their preferred input and output language, allowing questions to be asked in one language and responses delivered in another. This means users can interact with the Digital Workforce in their preferred language even where the underlying JD Edwards environment is configured differently.