Early Wins From Testing Our Own AI in Finance

In this article we discuss:

• Why Finance Needs a Customer Zero Program

• Early Wins From Testing Our Own AI

• 3 Lessons Leaders Can Learn From

• What’s Next

Finance is at a critical pivot point. Organizations that re-architect their processes around AI, rather than layering it on top of what exists today, will stand out from the rest. At Workday, we’re helping our customers seize this moment by acting as Customer Zero for agentic finance. This means we’re the first to use Workday’s finance products in order to build the path for others to follow.

Why Finance Needs a Customer Zero Program

One of the things I find most exciting about my role at Workday is sitting at the intersection of product and finance. 

Not only does my team drive transformation within our own finance organization, we partner hand in hand with product teams to make sure the features and capabilities they release are enterprise-ready. This means everything launched by Workday is built to scale and built to be trusted. We get to see some cool things first, but more importantly, we make sure there's trust and security in the products our customers receive.

That bar for trust is especially high in finance because a 5% error margin isn't just inconvenient; it's a regulatory risk. If you’re off by 5%, your CEO or CFO could face criminal liability.

This is why being Customer Zero can’t mean testing in a lab. We stress-test our products against the operational realities of a global enterprise, so our customers inherit a proven, safe playbook for scale. We're the guinea pigs, but that also means we get to try our technology early and build agentic finance into the way we work, fundamentally changing how we operate.

Beyond testing, we partner with our product teams to maximize AI and feature value by identifying new use cases and enabling stakeholder adoption.

Early Wins From Testing Our Own AI

Three AI products that have already been through our Customer Zero stress test, and the results are extremely promising.

Contract Intelligence: Reclaiming Millions Annually

The agent we get the most use out of today is Workday Contract Intelligence, powered by Evisort AI. It has already saved us tens of thousands of hours and helped us avoid millions of dollars in costs.

We’re using Contract Intelligence to reduce the time Workday employees spend scanning contracts, looking for terms, tracking renewals, and spotting opportunities to save on rates. It scans all of our contracts and surfaces those insights up front, so people can focus on making the decision rather than hunting for information.

Sourcing and procurement teams are also using Contract Intelligence to prepare for contract renewals. It gives them a quick view of previously negotiated terms across master services agreements (MSAs), statements of work (SOWs), and amendments. With that context, sourcing managers can go into negotiations better prepared, extend existing terms and discounts, and pull other levers to maximize savings.

Once we saw the time and cost savings in procurement, we asked, "What else can it do?" We are working with various teams that manage contracts at Workday, like real estate, to find opportunities to standardize, negotiate better terms, and build clauses specific to Workday. 

We aren't stopping at traditional use cases. We are looking at use cases that leverage best in class capabilities of Evisort, like optical character recognition (OCR). We realized we can use OCR to identify value-added tax (VAT) and goods and services tax (GST), ensure compliance with local regulations in different countries, and spot opportunities to recover taxes. This simple use case opens the potential to reclaim millions of dollars in tax leakage that would previously have been paid out. 

This is an example of how an existing AI use case can be expanded to drive value across the finance function. We’ve shared our success back with Workday’s product team, so it can help customers find new value in our tools as well.

Sana Enterprise: Upskilling Finance Practitioners to Be AI Builders

Sana Enterprise has proven to be a game-changer for my team. It turns every finance professional into a citizen developer, putting AI technology directly in their hands to not just operate but build.

With Sana Enterprise, anyone can create their own agents, reshape workflows, and build tools that augment their work. One of the biggest results we’re seeing is time back to be more strategic and apply human judgment, becoming true thought partners for the business.

Finance teams are underwater. Month-end close comes every month, quarter-end close every quarter, and the operational work never stops. With Sana Enterprise giving time back, it fundamentally changes what finance can be and elevates the entire function. It’s something the leaders I speak with are genuinely excited about.

Sana Enterprise has also cut our business requirements gathering from weeks to days. Sana Meeting Recorder captures stakeholder conversations, our agents turn them into business requirements documents, and Sana’s connectors link everything to our process blueprints in Miro and our process documentation. Review and sign-off cycles are shorter too, because each round builds on what we learned in the last one, so we get better at translating business needs every time.

The same is true for user acceptance testing, from writing test scripts to reviewing them. With that speed, finance can keep up as business strategy changes without being a bottleneck to growth.

Agent System of Record: Balancing Control Without Slowing Innovation

AI innovation is moving incredibly fast, and agents are popping up everywhere. A lot of leaders, especially CIOs, are asking: How do we orchestrate all of this? What happens when something goes wrong?

As Customer Zero, we use Workday’s Agent System of Record to create that orchestration layer. Instead of unknown agents operating on their own, employees register their agents in one place. 

What’s been most useful for my team is that anyone can see which agents already exist, so no one rebuilds something that's already there. That saves time, and it gives us control from an enterprise risk perspective. If something goes wrong, there's a central repository where we can take action.

We want innovation to happen—that's how you move with velocity. But having that control and orchestration layer in place creates a blanket of security, which has been a critical lesson for Workday as we empower more people to build their own agents. 

3 Lessons Leaders Can Learn From

To me, Customer Zero doesn’t just mean going first—it means we're building the path others can follow. Here are three lessons leaders can take away from our experience:

1. Re-Engineer Your Processes First

There’s real danger in layering AI on top of legacy processes. AI won't stop to ask whether the process underneath is wrong. Put AI on top of something broken, and you get broken results. 

Many organizations wonder why they aren't getting value from AI, and often it's because they haven't re-architected the processes it runs on.

That’s why we aren't using AI to pave over broken processes or chase isolated quick wins at Workday. Instead, we’re re-engineering finance from the ground up into a governed, scalable engine to pave the way for our agentic future.

Looking ahead three to five years, finance organizations that take the time to embed AI into the foundation of their processes will set themselves apart from those that simply layered AI on top of what they already had.

Related Read: Finance Teams With AI Deeply Embedded Are Pulling Ahead. Here's Why.

2. Build an AI-Ready Data Foundation

One of the biggest barriers to scaling AI beyond pilots is data. Pilots usually run in safe environments on a curated data extract. Once you move into production, you depend on the quality of data coming from upstream systems. That's why the data foundation is such a critical first step. Getting the data right is what lets AI scale, run in production, and earn trust, which matters enormously for a finance organization.

That means eliminating duplicate data and making sure data is properly governed. Financial systems sit downstream from almost everything else: sales contracts and terms, start and end dates, discounts, and expected revenue all flow into finance. It starts with understanding all the systems that feed process execution and identifying the critical data objects within each system that the process depends on. Without the right controls, a change in a sales system can flow downstream and leave you unsure which data is correct. Putting controls in place upstream ensures high-quality, compliant data reaches your financial systems, creating the strong foundation you need to automate processes and build agents that solve problems instead of hallucinating.

Related Read: AI Exposes Gaps in Financial Data Infrastructure

3. Bring Your People Along

The human element is often an afterthought, but it's one of the most critical parts of getting real value from AI. We can build the shiniest, coolest AI product, but if we don't bring teams along on the journey, it doesn't matter how amazing it is.

That's why we've launched programs at Workday like AI Ambassadors and Transformation Pulse. Our AI ambassadors have been critical in identifying new use cases for our agentic capabilities—resulting in expanding Contract Intelligence to corporate insurance contracts and using Decision Intelligence-Accounting in tax and treasury.

We use these councils and updates to showcase the great work people are doing to get us across the finish line, and to show others that transformation doesn't have to be hard. It can happen one step at a time. 

As leaders, we play a key role in making adoption manageable while keeping up the pace. At Workday, one way we do this is through persona-based narratives that help users better understand how processes are being reinvented and their specific role in them. We aim to showcase how Workday employees can become more strategic by shifting to value-driven activities.

Everyone plays a part in a successful AI transformation. It can't be done to people; it has to be done with them. 

Related Read: Why Being a More Strategic CFO Starts With Culture, Not Tech

What's Next: The Road to Lights-Out Finance

I get most excited about the Workday solutions our team is going to adopt next. Next up are Workday’s Financial Audit Agent, the Financial Test Suite, and new capabilities across financial close. They'll fundamentally change how we operate finance, drive significant savings, and move us meaningfully closer to a zero-day close and lights-out finance—a finance operation where routine work runs automatically and our team focuses on exceptions and judgment calls.

And we're only at the tip of the iceberg. As we strengthen our controls and foundation, and start building agents on top of agents, we'll unlock capabilities that would have seemed impossible even a year ago. As Customer Zero, we'll keep proving each new product on our own books, so our customers can follow with confidence.