WORK

What an AI-Enabled HR Department Looks Like (in a mid-sized distribution or manufacturing company)

A brief note: This article first appeared in ASA Review Magazine, but I wanted to share it here, as many of you reading work in frontline companies. Though the scenario below is theoretical, AI-enabled HR departments are very real. However, just because you can, doesn't mean you should. Read more about that at the end of the newsletter!

If your HR department were running on AI, the technology is not what you would notice.

You would notice the warehouse role that used to sit open for two months getting filled in two weeks. The new hire who raved about her first day instead of spending it hunting for a login. The branch manager who found out someone was thinking about leaving while there was still time to do something about it.

The transformation is quiet. It shows up in the flow of work, not on a dashboard. And almost nobody in our industry has seen it yet.

That last part is the opportunity. Last fall, Gartner surveyed HR leaders and found 88 percent had not seen significant business value from their AI tools. Only 12 percent had. SHRM's State of AI in HR 2026 report tells a similar story: just 35 percent of midsized companies, the range where most distributors live, have implemented AI in HR at all. (Among the largest enterprise companies, it is 60 percent.)

Put those numbers together and they tell us that most of us have not started. Most of the ones who did are not getting value yet. So the advantage is still there, and it will go to the companies with the best approach, not the biggest budget.

Now the number that should keep you up at night. A 2026 BlackFog survey found that 49 percent of workers admit to using AI tools their employer never approved, many of them on free versions, feeding in sensitive company information. Customer details. Employee records. Pricing. And in a separate survey from Okta, nearly all executives said they were confident their people were using AI responsibly. Their own workers told a different story. AI is already in your building. The only question is whether it is governed.

So how do you get from here to there? In my work with distributors, I see four levels of people strategy, and companies climb them one at a time.

The first is paper-dependent. Everything lives in a filing cabinet and a manager's head. The second is digital-basic, where you have systems, but they do not talk to each other, so your people spend their days moving the same information from one screen to another. The third is integrated-smart. Your systems finally connect, data flows, and you can actually see what is happening across the business. The fourth is AI-enabled, where the work starts anticipating you. Roles get filled before they drain a branch. Risks surface before they walk out the door.

Most distributors sit between the first two levels, running on spreadsheets and disconnected systems. That is not a failure. That is just where our industry is. And you can’t expect to jump straight from spreadsheets to magic. Like any change effort, it takes time and patience to get there. But you will get there. Let me walk you through a random Tuesday for a midsized distributor with an AI-enabled HR department.

6:00am. A supervisor opens a branch and needs to know how holiday pay works for someone who started mid-year. He does not wait for HR to check email. He asks an assistant trained on your own handbook and gets the answer in the parking lot. If your company runs locations in six states, as one of mine did, that assistant knows which state he is standing in. Sick leave in one state, not the next. Different overtime thresholds. Different final paycheck rules when someone quits on a Friday. That knowledge used to live in the head of whoever had been in HR the longest, and when she was on vacation, the branch guessed. Now it does not guess, it knows.

7:15am. Two drivers call out. The branch manager used to spend the next forty minutes on the phone working a puzzle: who is qualified, who is under hours, who has already hit overtime this week, who said yes the last three times and is starting to resent it. Now he opens his phone to a ranked list of who can cover, already filtered for qualification and hours, with the overtime cost of each option shown next to the name. He still makes the call. But he makes it in four minutes instead of forty, and he can see that the same two people have absorbed every callout this month, which is a retention problem waiting to happen.

8:30am. Your HR manager opens her day. Three applications came in overnight for the open driver role, already screened against what has made your best drivers succeed. She has a short list to weigh, not a hundred resumes to sort. And here is what is not in her inbox: the employment verification for someone's mortgage, the PTO balance question, the sixth request this week for a copy of the benefits summary. During open enrollment, those questions used to arrive by the dozen, every one of them answerable from a document she had already written. They are handled. She has not stopped being available to people. She has stopped being the only place someone could go with a question.

10:00am. A flag comes up on one of the best warehouse associates in the building. Nothing dramatic. She has stopped picking up the extra shifts she used to take. Her scan rate has drifted. She called out on a Monday, which she never does. None of those things means anything alone, and no human tracks all three at once across 700 people. Together they are the same pattern that showed up in the last dozen people who left. It surfaces two to three months before a resignation letter would. So her manager buys her a coffee and asks how things are going, and finds out she has been picking up a second job because her hours got cut in the spring. That is a fixable problem, but only if you find out in July instead of September.

11:30am. The system is wrong about someone. It flags a long-tenured guy at another branch as a turnover risk, and his manager knows exactly why the pattern looks strange: his mother has been in and out of the hospital since May. The manager overrides it, notes it, and moves on. This happens. It will happen at your company too. AI-enabled tools are just that, and the value only shows up when there is a person close enough to the work to know what is missing. The companies that get burned are the ones that treat the flag as the answer, then act on it, then wonder why their best people feel like Big Brother is watching them. Nobody gets hired, promoted, disciplined, or let go by an algorithm. The algorithm points. The human decides.

2:00pm. The new driver hired two weeks ago is out on a route. Her first day did not involve a clipboard or a scavenger hunt for a login. When I map onboarding with clients, we routinely find 15 to 20 manual steps between the offer letter and a productive first week. Done well, with a person still in the loop, that becomes maybe three. The rest happened quietly before she ever walked in the door, which is why she spent day one learning the job instead of waiting on paperwork.

That is the department. Not a robot in the corner. A branch that answers its own questions, a schedule that solves itself before lunch, an HR manager doing work only a person can do, and a leader who found out in time.

Now notice who that Tuesday belonged to. The supervisor in the parking lot. The branch manager working the callout. The associate whose hours got cut. Not one of them sits at a desk, and that is exactly who most of this technology is not built for. Eighty percent of the global workforce does not work at a desk. BCG's 2025 AI at Work study found 75 percent of leaders use generative AI regularly, compared with just 51 percent of frontline staff. BCG calls that gap the silicon ceiling. It is not about willingness. Enterprise AI is built for desktops, not shared handhelds. It assumes someone has time to stop and type a prompt. It assumes connectivity a lot of warehouse floors do not have. So most of what you will be pitched this year will work beautifully for your office and never reach your floor. If that is your plan, you have automated your headquarters and abandoned your warehouse.

And think about what changed for the HR manager, because that is where this reaches the bottom line. The typical midsized distributor runs one HR person for every 100 to 200 employees. When I have HR audiences track their week, transactional work eats about 36 hours of it. Four hours left for anything strategic, if nothing goes wrong, and it always goes wrong. The verification letters and the PTO questions that left her inbox are not a small convenience. They are the difference between an HR department that processes your company and one that builds it. Those hours go to coaching, to retention, to figuring out who you will need next year. No new headcount required.

So how do you start? Not by buying anything. Read the failure stories behind that 88 percent and the pattern never changes: they bought the technology first and went looking for the problem it solved later. Do the opposite. This week, name your single biggest people-related time drain. This month, point one approved tool at that one problem, set a baseline so you will know if it worked, and tell your team exactly what it does and does not do. This quarter, expand what worked. One tool, one problem, one month. That is the whole entry fee.

Your people strategy is your business strategy. AI has not changed that. It has only changed how fast the gap widens between the companies that act and the ones that wait.

COMMUNITY

Left to my own devices, I would spend most of my weekends at home, visiting local coffee shops, and be very happy! I don't need to be surrounded by a lot of people all the time, or really ever! However, it is good for me to get out in the world, and I actually love spending time in small groups of people.

I spent this weekend in Chicago and spent a lot of time with Alex Seiler, and we got together for brunch on Saturday with our mutual friends Brittany Storie and Angela Howard. Angela is Chicago-based like Alex, and Brittany was in town visiting, like me.

I've known Angela for years, but I met Brittany more recently. However, none of that mattered as we spent hours talking life, work and everything in between.

Community has taken many different forms for me over the years. From really not much early in my career, to large HR communities, to creating my own. And now, I really find that I love spending time in these small groups, microcommunities within larger communities of people I adore.

Alex, Brittany and Angela are all amazing humans, and absolutely brilliant people strategists. If you aren't already connected on LinkedIn, click the links above!

COFFEE

I will walk, drive or take any form of transportation for great coffee or a cool coffee experience.

I walked over a mile in tropical humidity to visit Blue Bottle in Chicago this weekend. (Yes, I know it's a chain. But I love it, and the closest one to me is Boston!)

I realized when I was there I was less than a five minute walk from the Starbucks Reserve, which I try to go to every other visit to Chicago or so. I was due for a visit, so I went.

If you haven't been, it is quite an experience. It was also absolutely packed at 7:45AM - I did not expect that. But, it has coffee and merch you can't find anywhere else, so for coffee fans like me, it was definitely worth the trip.

And a little bit more.......

Last time I shared my updated website, below. I have a few more things coming, so definitely watch that space!

For now, though, I wanted to go back to what I shared with you about AI, because I have a lot of things coming up related.

In October, I'll be again teaching a course for UNH as part of their HR Certification Program, on AI in HR. This is an AI literacy course, created for those who are still new, though everyone will find something interesting in it. I'll also be co-facilitating the second cohort of an AI fluency program with Drew Fortin at Lever Talent. This was really fun to do, and is meant for people a little further on in their journey. Still in October, I'll be closing out the HRLF event with a keynote that's not on AI, but we're doing something really unique and cool (and secret) that is AI-based, at the very end of the event. And then, the following week, I'll be co-presenting at a Paylocity event at Gillette Stadium with Brian Bouchard of Sheehan Phinney, "Just Because You Can, Doesn't Mean You Should", all about AI.

I have lots of non-AI things going on too. I'll share more info on these as we get closer but for now, if you want more info, just reply to this email!

Have a great week! See you soon.

Tracie