Q: What excites you most about Oracle’s Quote-to-Cash direction? 

A: Oracle has a clear vision for bringing more of Quote-to-Cash onto one platform. Other providers may offer important pieces of the lifecycle, but they frequently require additional applications or third-party technologies to create a complete enterprise solution.  

Historically, Oracle CPQ handled configuration and quoting while the “to-cash” portion of the Q2C process typically lived within ERP. Depending on the customer, that ERP might have been Oracle E-Business Suite, Oracle Fusion, JD Edwards, SAP or another system.  

Oracle is continuing to bring enterprise-grade quoting and downstream revenue processes closer together within the Fusion platform and its unified data model. That makes the company’s end-to-end Quote-to-Cash story much more substantive than it was when CPQ and ERP remained more visibly separated.  

The important point is that Oracle is building for complex global enterprises—not just businesses with straightforward selling models. The broader suite is increasingly capable of supporting intricate product, pricing and organizational requirements without forcing customers to assemble numerous add-ons or extensively customize every area. 

“Oracle’s advantage is the breadth of the vision—from quote through cash—on a more unified foundation.” 

Q: Why does the broader system architecture matter so much? 

A: Enterprise architectures rarely result from one companywide technology decision. They evolve through acquisitions, regional requirements, and various best-of-breed investments made over many years.  

That is why many large companies operate with three or four major platforms participating in Quote-to-Cash. The applications may exchange transactions effectively, but the company does not necessarily have one unified source of data across the entire revenue lifecycle.  

That distinction becomes more important when an organization begins introducing AI. To train, ground and control AI-enabled capabilities, the business needs a reliable source of relevant data. When that information is dispersed across platforms, the company may need to consolidate it within a separate data lake or introduce another layer above its applications.  

The key questions become: Where does the data reside? Is it accurate? And can the organization use it confidently? Those considerations will determine whether a company can rely primarily on the AI embedded within one platform or whether it needs a broader, platform-independent approach. 

Q: What does “agentic sales” mean in a practical Oracle Q2C context? 

A: The most promising near-term opportunities are not autonomous agents independently creating highly complex quotes. Many enterprises do not yet have sufficiently clean data or configuration logic to trust that approach.  

A more immediately valuable application is using an agent to make established analysis easier for sellers to access. 

For example, a company may have a decade of transaction history showing win rates, pricing, and discount behavior across similar deals. Analytics can identify the pricing ranges that have historically performed best for a particular type of opportunity.  

An agent can then present that intelligence conversationally within the seller’s workflow. A salesperson could ask where a deal should be priced and receive a recommendation supported by relevant historical performance rather than having to find and interpret the underlying reports independently.  

The recommendation can also incorporate approval thresholds and commercial policies. The objective is not simply to suggest the deepest possible discount in pursuit of a win; it is to recommend a defensible price within the company’s rules and margin expectations.  

In that sense, an agent serves as an accessible layer that sits over data analytics, established policies and trusted business logic.  

Oracle’s product direction supports this interpretation. Oracle CPQ 25C introduced an AI-powered quote assistance agent and AI-generated product recommendations. Oracle’s 26C will include policy-aware deal-pricing guidance and natural-language configuration assistance for CPQ running in Fusion. 

“The agent should not replace the logic beneath CPQ. It should make trusted guidance easier for the seller to use.” 

Q: Can natural-language AI reliably configure complex manufacturing products? 

A: It depends on the product and the quality of the underlying configuration model. 

For a relatively simple transaction, asking an assistant to add several known SKUs may save a seller from clicking through multiple screens. That can make the experience faster or more convenient.  

Complex manufacturing configurations are different. If the system must interpret extensive dependencies, compatibility requirements, and customer-specific rules, the necessary product data and configuration logic must already be accurate and complete.  

A new conversational interface does not eliminate the need for disciplined product modeling. It simply provides another way to access the logic and information underneath it. 

Q: Is AI changing the case for unified enterprise platforms? 

A: Enterprise technology strategy has historically moved back and forth between unified suites and best-of-breed applications. 

As integration technology improved, companies gained more freedom to select the strongest application for each area and connect those applications through middleware. The convenience of having everything on one platform was not always enough to outweigh the perceived advantages of best-of-breed solutions.  

AI may shift that calculation. 

When more of an organization’s relevant data resides within a common data model, it can be easier to analyze and use as the foundation for AI-enabled capabilities. A company may not need to build and maintain a separate repository simply to consolidate information from every application before it can put analytics or agents to work.  

That is one reason Oracle’s unified platform vision is compelling. But the high-level direction aligns with the growing need for connected, AI-ready data.  

Q: What should Oracle customers be doing now to prepare for agentic sales? 

A: It’s all about the data. Determine whether your product, pricing, customer, and transaction data is accurate and can support the technologies you want to implement. If cleanup is required, do not wait for a transformation project to begin. Data remediation is frequently the longest part of the effort, so the right time to start is now. 

A majority of our clients need data cleanup before AI will provide the business value organizations are looking for.  We are working with those clients to define what data changes are required and preparing an actionable plan to help guide them through that effort.   

At the end of the day, preparing for agentic sales requires the same fundamentals: reliable data, trusted business logic, and connected processes. Agentic sales will change how sellers interact with Quote-to-Cash. But the impact and results still depend on the foundation beneath it. 

Meet the GM: Dave Farley, General Manager, Pierce Washington Oracle Practice 

Dave is a University of Notre Dame graduate who began his career in Chicago before moving to the Bay Area with Oracle and later returning to his native Minneapolis. Dave’s career has included leadership roles in Oracle CPQ consulting, operations and product management. He previously served as a Principal Product Manager and Consulting Practice Manager at Oracle. Before that, he was a solution architect and business analyst at BigMachines and an analyst at Accenture. Today, Dave is General Manager of Pierce Washington’s Oracle Practice.