Here Comes “Workday AI”
Today at the Rising conference Workday introduced a range of new generative AI features, all described as “Workday AI.” This is part of a major trend: enterprise software companies adding Gen AI features to their products.
The first message from Workday is that the company has been doing AI and machine learning for quite some time. Going back to the acquisition of Identified in 2014, there has been AI and ML in the system for years. Second, Workday is now also committed to Generative AI, and they want to do it in a “platform-centric, trusted way.” All great for Workday customers.
What they introduced, which I consider first Gen AI features, was quite extensive. For example, you can automatically create job descriptions from your Workday data set (a nice feature but not at all unique to Workday) and customize them based on various job parameters. You can automatically create help guides and “how-to” docs from Workday documentation, enabling users to ask questions about policies and how-to issues. And in the finance area you can automatically analyze purchase orders and other contracts against related transactions for consistency, letting the AI find inconsistencies.
These are important and useful features customers will enjoy. We don’t really know how what LLM Workday is using (ie. how the system is trained), but Workday assures us we can trust the system because the LLM uses “clean, reliable Workday data.” And given Workday’s 60 million users (some of whom have opted in to let their data be used for training), there is a lot of intelligence here. So throughout Workday we’ll see more and more built-in generative features.
If you think about all your corporate applications (HR, financial, marketing, etc), this kind of functionality will start to appear everywhere. These systems have lots of information about your company (all your existing job titles, levels, and descriptions, for example), so something like writing a job description can suddenly become easy and automatically consistent with other data in the company.
(It does make one wonder when we will want to use the Microsoft Office copilot vs. Workday, but we can explore that later.)
On a more advanced level, Workday also introduced a system called Manager Insights Hub which I found very impressive. This is a completely integrated Manager portal that lets any supervisor or leader look at employee goals, development plans, feedback (integrated from Peakon), and other information in a single integrated page. This new system is designed to make managers “better at managing,” a problem every company faces. And to make it even snazzier, there is a new Gen AI feature that creates employee “Growth Plans.”
What is a “Growth Plan?” It sounds like a nicely worded performance plan, focused on development. The AI “quickly creates a summary of employee’s strengths and area of growth, pulling from stored data including performance reviews, feedback, contribution goals, skills, sentiment, and more.” And the demo showed that each input has a source, so the employee and manager can see “who said what.”
This is a really creative idea but also a bit of a risk. While Workday shows the user where the feedback came from, it may come from email exchanges or poorly written performance reviews, leading an employee to dislike or possibly mistrust the result. Products like Textio Lift (the leader at this) use Gen AI and decades of research to make sure feedback and performance comments are unbiased, actionable, and legal. Workday will have to make sure their system does the same.
And now that Workday AI is here, there are lots of Gen AI opportunities yet to come: automatic search and indexing of learning, generative AI to build courses and assessments, and narrative tools for analytics, reporting, and cross-domain analysis. Now that the “Workday AI” moniker is live, we can expect more to come over time.
But there’s more: the super-duper “Ask Workday” assistant that Jeff Gelfuso demonstrated as a work in process. This feature, which I’ve discussed with the team for a few quarters, is the “big one” Workday customers really want. And it feels like SAP beat them to the punch with Joule (announced yesterday).
This capability, which will let users ask Workday questions like “how’s our hiring going?” or “what’s a good career path for me?” is really where Gen AI in HR is trying to head. Employees, managers, and HR teams want to bypass the Workday menus and screens and get to the information and transactions faster.
Based on what Jeff demonstrated, I’d guess this will arrive in the coming year. So before long we’ll have a Workday Copilot that I know Workday customers will love. (The SAP Joule product has been in development for a while and it will debut at SuccessConnect next week).
My last quick note is the Workday AI Marketplace and Workday AI Gateway. These are really interesting ideas, enabling customers, developers, and integration partners to integrate their AI tools and build new AI models that access the Workday data set. We can imagine dozens of creative solutions to appear here, so this is also a big play to help Workday expand its AI capabilities even faster.
What This All Means
It’s clear from this announcement that Workday takes Generative AI seriously. While the company didn’t really expose its underlying tech stack, these announcements show customers that Workday is focused here and is placing some educated bets. And it’s also clear that Workday wants customers to trust these features since they’re essentially mining a company’s entire HR and Finance history for analysis.
What’s still missing? There really isn’t a talent intelligence strategy yet (SAP Talent Intelligence Hub comes out this quarter), the generative features are still in limited places, and the end-to-end AI assistant is in the works. But Workday customers tend to be conservative companies and they use the system for global, mission-critical applications, so Workday’s focus on “platform-integrated AI” makes sense.
I will write more next week as some other important announcements hit the wire. Stay tuned, the next few weeks are important times in HR Tech, we’ll be seeing many new AI Copilots and more throughout the month of October.