How to Transition Your HR Department Into the AI Era | by Nikos Rigas | Jun, 2026 | MediumSitemapOpen in appSign up<br>Sign in

Medium Logo

Get app<br>Write

Search

Sign up<br>Sign in

How to Transition Your HR Department Into the AI Era

Nikos Rigas

10 min read·<br>Just now

Listen

Share

Most companies bought AI for everyone and got little back. Here’s how to build one that actually makes a department more productive.<br>Press enter or click to view image in full size

I’m Nikos, Head of Engineering at Cenobe. I spend my days designing systems that have to run at scale — and just as much time watching how real organizations actually work: where they move fast, and where they quietly get stuck.<br>This article isn’t a sales pitch, and it isn’t a list of tools. It’s a mindset. I want to walk you through how to genuinely move your HR department into the AI era and make it more productive — not just more expensive. Because right now, most companies are spending real money on AI and getting very little back.<br>So let’s start with where we are.<br>Where We Are Now<br>Before we talk about HR specifically, it’s worth being honest about the state of AI at work today. For most companies, it hasn’t gone the way they hoped — and it’s worth understanding why before spending another euro.<br>The “Buy Everyone a Seat” Mistake<br>For the last couple of years, almost every company I talk to has poured its energy into the same thing: putting AI into the products they sell. A copilot in the app. A chatbot on the website. A “powered by AI” line on the pricing page.<br>Meanwhile, the people who actually run those companies — the departments that keep them alive — got handed a license and a login.<br>That was the whole plan for “AI transformation” at a lot of organizations. Buy everyone a seat to a general-purpose assistant, send a launch email, and wait for productivity to climb. A year later, the climb hasn’t come, and the finance team has noticed. The bill is real; the gains are vague.<br>And this isn’t just a feeling. In 2025, MIT looked at how AI was actually being used inside companies and found that about 95% of these projects made no real difference to the bottom line. Gartner studied Microsoft’s Copilot specifically and found that most companies simply couldn’t point to a clear payoff for what they were spending. And they were spending — around $30 per person, every month. Pay that across a whole company, see little come back, and sooner or later someone in finance starts asking hard questions.<br>So executives are doing the natural thing: getting grumpy about AI spend. And I think they’re asking the wrong question. The question isn’t “was AI a mistake?” It’s “why did we expect a generic tool to transform a specific department?”<br>Why Generic AI Falls Flat<br>MIT also looked at why this kept happening, and the answer wasn’t “the AI isn’t smart enough.” The models are good. The real problem is that a general-purpose AI tool doesn’t know anything about how your team actually works — and it never learns. In fact, when companies built AI around a specific job instead of buying a one-size-fits-all tool, it worked about twice as often.<br>Almost everyone skipped two things.<br>The first is training the people. You can’t hand someone a blank chat box and expect the way they work to change. Most people, given no guidance, use it to tidy up the odd email and then forget it’s there.<br>The second — and this is the big one — is making it fit. Every department has its own rules, its own documents, its own repetitive work. A generic assistant knows none of that. It’s never read your policies, your contracts, or the decisions you made last year. It’s like a brilliant new hire on their first day — except it stays on its first day forever.<br>Buying a tool isn’t the same as helping a department change the way it works. If you want a part of your business to truly run differently, you have to build something that actually knows how that part of the business runs. Let me show you what that looks like for the department where it pays off fastest: HR.<br>What an AI-Era HR Department Looks Like<br>Let me start with the destination. Picture the people inside an HR department a few months after building AI around the way they actually work — not bolted on, but shaped to fit.<br>Someone on the team gets a tricky question — say, an employee wants to know if they still get their bonus while on parental leave. Normally she’d have to dig through a stack of policy documents, or stop a busy senior colleague to ask. Instead, she just asks the team’s assistant. In a few seconds she has a clear answer, taken straight from the company’s own rules, with a link to the exact document it came from. She replies in minutes, sure she’s got it right.<br>Across the team, knowledge stops being siloed. The person who knows immigration cases inside out, the one who’s handled every tricky termination, the one who wrote the relocation policy — what they know is no longer locked in their heads and their inboxes. It sits in one place...