One of my first jobs when I started at Workday was to see if we could not only implement for ourselves the AI suite that we were selling to customers, but also measure the results. We could. In the first nine months of embedding AI into our end-to-end hiring process at Workday, my team got back 24,000 hours — time spent on manual resume screening and interview scheduling — and we increased recruiter capacity by roughly 40%.

I talk to a lot of HR and talent leaders, and when I share that figure, I get three reactions, pretty much in this order — shock at the size, then, 

You're the first people we've talked to who've actually seen ROI from AI,

 and, almost cautiously, 

Wait, you didn't reduce headcount?

We didn't. But we did make changes. And what it took to make that work, evolve the team, and protect a sensitive domain like hiring was new, strategic, and intentional.

For most of recruiting's history, a big chunk of the job was administrative. Clearing the queue. Screening hundreds of applications, coordinating five calendars, nudging candidates from one stage to the next so the process didn't stall. It all was necessary, but let's be honest, nobody got into talent acquisition because they love pushing paper. Most of us got into this line of work to talk to people and help place them in great roles.

So when we recovered those hours, the question was what to put in their place. For us, the answer was to focus on the work recruiters like most - sourcing harder-to-find candidates, prepping people properly for interviews, coaching hiring managers, building relationships. We've started calling it “talent orchestration” when the administrative weight comes off, and a more human version of the job grows in the space it leaves behind. If all you do is make the old admin faster, you've missed the point and the opportunity.

I want to be transparent here. We did not flip a switch and harvest 24,000 hours. And AI doesn't replace all jobs.

The reason we could move fast on AI is because Workday had spent the prior couple of years acquiring the right pieces — HiredScore for AI-driven ranking and matching, Paradox for conversational scheduling and frontline hiring, Sana for a conversational front-end to work. That foresight from our product and technology leaders is what let us treat AI as a layer we could strategically stack on a platform our employees and our customers already trust, instead of running it as a side project to see if it worked.

Even with the right tools, the change was hard. Recruiters had spent careers building muscle memory around the manual steps. Asking them to trust an agent to rank a candidate pool, or to let conversational AI schedule an interview in seconds instead of three days of phone tag, ran headlong into established instincts.

But we kept adjusting. The work didn't disappear so much as rebalance, and it rebalanced almost week by week. We realized we suddenly needed someone who could build and maintain an agent or own a piece of the candidate experience that hadn't existed before. So we repurposed. We re-skilled. We treated the time we'd recovered as capacity to redeploy rather than reduce headcount.

A year ago, the fear of AI taking your job as a recruiter was real. Anxiety in the market was running high, and plenty of companies were releasing headcount and blaming AI. But six months in, my team came to understand that AI was augmenting their work, allowing them to hand off manual and tactical tasks, creating space to spend more time on the work that they like best.

Much of our routine HR work has been taken on by two agents. The Recruiting Agent handles sourcing, screening, and matching — comparing the skills named in a job description against the skills shown in an application and surfacing the strongest matches. This reasoning is transparent so anyone in our hiring team can see how the agent arrived at its recommendations. The Candidate Experience Agent manages the conversational, high-volume touchpoints - scheduling, reminders, answering the candidate's questions — all the things requiring speed and accuracy.

What I didn't anticipate was that the AI stack would continue to grow organically. People on my team started building their own agents on top of the AI we deployed. They've shipped a couple already. This new capability didn't exist before, and it's the clearest indication that we're rebalancing roles, not subtracting them.

This matters for the volume problem too. AI has made it easy for candidates to tailor a resume and apply at scale, so every employer, us included, is drowning in applications. Trusted systems that can make sense of the volume are how you stay fair and responsive when the front door is being flooded.

We use our own tools, which means I can speak to agentic HR as a practitioner, not just a vendor — and increasingly, so can our customers.

7-Eleven automated 95% of its hiring process with conversational AI, returned more than 40,000 hours per week to store leaders, and took time-to-hire from 10-plus days to less than three.

JLL, the global real estate firm, was fielding 1.5 million applications a year. After bringing in AI-driven screening, its time-to-fill dropped from 52 to 30 days, lifting quarterly hiring 64% — from 3,180 hires to 5,125. Recruiters who could only manage 10 requisitions at a time could now tackle 50 concurrently. The team went from playing catch-up to advising the business.

AdventHealth, in the middle of a nursing shortage, used AI to resurface qualified candidates sitting in its database that its recruiters had never had time to revisit. It was able to hire roughly 300 nurses, double requisition close rates in 90 days, and reduce hiring manager decision time by 40%.

Among our customer base exceeding 11,000 businesses, more than 75% are now using Workday AI in their core processes, and more than 2,000 are running AI agents in general availability.

We reinvested every hour we got back into making recruitment more human — more time with candidates, more time with hiring managers, more time building capabilities that didn't exist on my team a year ago. We're not doing the same hiring faster. The job itself is becoming a better one, and that's only true because we were willing to change our habits, govern the high-risk parts with discipline, and treat the recovered time as something to invest rather than pocket.

Disclosure - the author is an employee of Workday. Figures cited reflect Workday's own implementation and published customer data.