Our ETR Insights team interviewed a veteran CIO in the financial industry who reviewed AI integration across enterprise platforms such as Salesforce, ServiceNow, and Workday. During the interview, they note a shift from general enthusiasm for AI, towards more disciplined, practical implementations focused on efficiency and profitability. Our guest is particularly familiar with Workday's Illuminate, which they describe as already transforming HR and finance processes through multi-agent orchestration. However, they believe that legacy system integration and data quality remain sticking points, and that ROI is not always easy to measure. Competition from Microsoft Viva, Oracle HCM, Eightfold AI, and Visier continues to intensify, with each offering distinct strengths and trade-offs; the interview touches on shifting organizational models, specifically Workday's transition toward skill-based roles and ServiceNow's push for cross-departmental automation. Data stewardship programs, how enterprises leverage historical data, batch processing, and real-time dashboards, and about governance frameworks for AI within HR are also discussed.

Vendors Mentioned: Eightfold AI, Microsoft (Power BI, Viva Insights), Oracle (HCM Cloud Analytics), Salesforce, ServiceNow, Visier, Workday (Illuminate)

HCM: From ‘Wow!’ to ‘How?’

Enterprise software providers, including Salesforce, ServiceNow, and Workday, are rapidly integrating AI. “I would say we moved from, ‘Wow!’ to ‘How?’” says our guest. “We moved from mid-early 2023, when the first AI solutions became better known to boards—and their willingness to implement AI—to a much more selective approach in terms of where AI makes sense.” What felt earlier a general enthusiasm for all things AI is now a more disciplined evaluation process, with boards asking not just where AI can be used, but more specifically how it can drive efficiency and profitability. “We [are moving] from AI which has been kind of understood as “gen AI” or “GPT” by everyone, to the next iteration of agentic AI, where AI can make different actions and choices based on the data and predefined parameters.”

Workday’s artificial intelligence initiative, Illuminate, uses both data and contextual insights to drive productivity and decision- making within HR and finance. Our guest offers detailed implementation experience. “This is a solution not only to automate manual tasks and assist employees,” they report, “but ultimately to transform entire business processes,” including anomaly detection, prompting, and document scanning. At the core of Illuminate is its orchestration functionality, which allows AI agents to operate across various business processes on behalf of users. “Basically every user has the ability to oversee, and AI agents can operate on behalf of users to orchestrate various business processes. I think this is probably one of the most interesting cross-platform, multi- agent deployments of Illuminate.”

That said, “Agentic AI systems are still new, and making autonomous decisions can sometimes happen in unpredictable ways.” Our guest warns, “Especially in business-critical systems like HR systems, it’s still quite risky.” Beyond reliability, integration with legacy systems presents another hurdle. “When you connect agentic AI with existing systems or legacy systems, it requires good API management, data flow orchestration, and the ability of some of these systems to work with the newer AI systems.” Data quality is of similar concern. “Illuminate’s machine learning capabilities are heavily dependent on clean, consistent HR data, which many companies struggle to maintain. I have seen data silos, data not being updated, and structure charts are not being up to date, especially as people move, leave, or join.” Companies need to strike a balance between automation and human judgment, and governance frameworks must establish clear oversight of when and how AI influences employment decisions, a process fraught with legal and regulatory implications. Our guest reminds us that with any new technology, enthusiasm for cultural change will vary. “You basically are making HR professionals trust AI-driven insights.” Finally, the matter of ROI: Measuring the impact of AI-powered HR analytics is no easy task. “Especially [for] long-term talent decisions such as succession planning, etc.”

Workday’s event-driven architecture demands sophisticated middleware to translate between systems. “The APIs with Workday infrastructure can have specific rate limits, authentication requirements, and data transformation needs that constrain agentic AI's functionality, especially if you need to have high-frequency operations.” IT teams must also manage disparate data sources. AI models rely on more than just internal HR data; they pull from performance metrics, market intelligence, and even employee communication patterns. “It requires data validation and cleaning in terms of implementing field-level validation at data entry points, so format checks, field checks, etc. [You have to create] cross-field validation rules, like the hire date of the employee cannot be before their birth date, etc. Deploying machine learning-based anomaly detection to flag impossible patterns or unusual patterns.”

Figure 1. ETR’s NOV24 AI Product Series asked Workday Illuminate users about the maximum price uplift their organization would be willing to pay for the vendor’s AI features (N = 40).

Traditional batch processing may be insufficient for AI agents that need current context. “As we move to real-time or near real-time, maintaining near real-time synchronization between systems becomes really crucial.” Agents also need historic data context to identify patterns, “But when you access the archive data with different storage methods or CMOS, it just adds to the integration complexity. That's talking about data integration challenges.” Companies are implementing data stewardship programs and real-time dashboards to monitor and improve data integrity. “When your data is incomplete or not diverse enough, then agentic AI will make wrong decisions. You need to look at, are there any missing data elements. What are the data collection campaigns' priorities for those missing attributes? What is the supplemental internal data, and also external datasets, in terms of industry benchmarks?”

No Shortage of Competing Tools

Our guest appreciates Workday’s hands-on implementation support. “One of the advantages I have seen personally with Workday is that their teams go above and beyond initial configuration or helping with the initial setup. They keep some of their implementation teams pretty much to the end of implementation and going live.” However, Workday faces increasing competition from well-known industry heavyweights. “Microsoft Viva Insights, which works with Power BI. I think it's a more communication and collaboration-focused alternative. It has excellent access into the Microsoft ecosystem, like Teams and Outlook, and very strong visualization through Power BI, though it has weaker functionality in structured HR analysis or formal talent workflows and processes.” Oracle HCM Cloud Analytics is strong in financial planning but trails in predictive workforce capabilities; Eightfold AI is an AI-native talent intelligence solution, purpose-built for talent acquisition and mobility use cases. “Another one is Visier, which is a best-of-breed people analytics specialist with more extensive pre-built analytics content libraries, vendor-agnostic, and very deep specialization in workforce planning, though perhaps it has less user experience in terms of usability compared to Workday.”

“Workday accelerates the shift from job-based to skill-based organizational models,” says this IT lead, describing more fluid internal mobility and a reduced reliance on middle management. “I think from an enterprise impact perspective, it drives centralization of people decisions. It enables local execution more, through improved visibility and automation, and creates centers of excellence around skills intelligence and talent marketplace operations.” By contrast, ServiceNow is positioning itself as an enterprise workflow orchestrator, pushing cross-departmental automation that breaks down silos. “The impact it has on organizational structures is that it enables true shared service models beyond traditional departmental boundaries.” Salesforce, meanwhile, is expanding beyond CRM into customer engagement, expanding their Customer 360 functionality across all customer touchpoints. “They’ve built sophisticated AI assistants for various customer-facing roles in sales, service, and support.”

Broadly: “You can see convergence trends reshaping companies, [with] three platforms building competitive employee experience functionalities. I think you need to watch out for confusion around, how do you govern and who owns those platforms, and how to you move from traditional HR and IT boundaries to experienced teams.”

Figure 2. ETR’s NOV24 AI Product Series asked Workday Illuminate users perceptions around the product’s Value vs. Cost (X-axis) compared with overall AI ROI vs. Initial Expectations (Y-axis; total N = 40).