Oracle AI World is happening next week, but as a preview, Oracle has a Fusion Cloud announcement up its sleeve. New AI agents are designed to help customer service, marketing teams, and sales departments.
Oracle’s agents are not an add-on to Fusion Cloud, but are included in existing subscriptions. This integration is characterized by the word “embedded,” which Oracle mentions several times in the announcement. According to EVP Applications Development Chris Leone, this integrated AI tooling should eliminate “reactive, manual and laborious” processes. The agents can perform such tasks autonomously. They do, however, report any deviations that may require more thought in order to make a decision.
Three categories of agents
The new agents are divided into three business disciplines. Three agents are available for marketing. The Account Product Fit Agent helps prioritize accounts that are likely to lead to purchases. It does this by combining predictive scores, account data, and engagement signals.
The Buying Group Definition Agent uses algorithms to identify industry- and product-specific purchasing roles. This should lead to more precise persona-based targeting. Finally, the Model Qualification Agent recommends the most suitable target groups and assesses the data quality for campaigns.
Sales teams will have access to six new agents. First up is the Deal Advisor Agent. This agent simplifies the preparation of product and price overviews, manuals, and customer references. The Quote Assistant Agent answers deal-related questions and streamlines quotation processes.
The Product Recommendations Agent then suggests cross-sell and upselling options based on customer history. The Quote Summaries Agent summarizes deal details, history, and next steps. The Contract Advisor Agent provides overviews of contract terms and conditions. The Lead Advisor Agent analyzes lead behavior and recommends follow-up actions.
In other words, the agents differ significantly in the complexity of the tasks they perform. However, their actual “agentic” nature unites them: where a person was previously needed in the process, agents have taken over the task from start to finish. It is up to physical staff to assess the practical results. The success of this depends both on Oracle’s agentic design and access to the data it has, which is a task for the end customer to arrange.
Customer service is being automated
Finally, six agents are also available for customer service. The Triage Agent analyzes service requests by product, category, severity, and sentiment to prioritize service. The Self-Service Agent acts as a customer-focused guide for information and problem solving.
The Service Request Creation Agent converts customer chats, phone calls, and emails into service tickets. Meanwhile, the Work Order Agent generates pre-filled draft orders. The Service Request Clustering Agent identifies recurring issues by analyzing similar requests. The Escalation Prediction Agent predicts which requests are likely to be escalated.
Read also: Okta weaves AI agents into an identity fabric