I’ll be at HR Tech 2026 in Las Vegas in two weeks, walking the expo floor the way I’ve walked hundreds of them over 25 years: looking for the handful of tools that actually change how we place people, and mentally filing away the rest as noise.

This year, I expect the pitch at almost every booth to include some version of “AI.” And after 25 years of matching candidates to employers based on more than keywords and automated screening, I’ve got a short list of things I need to see solved before any of it earns a place in our process.

I’m not looking for the newest technology. I’m looking for technology that can handle the reality of recruitment.

1. It has to survive contact with a real, messy candidate.

A lot of hiring technology is designed around the cleanest version of a candidate: a straightforward résumé, one language, and a linear career path. That is not who typically walks through our door.

It is the career changer with an unconventional title history, the immigrant professional whose credentials do not translate neatly to a Canadian equivalent, and the candidate who shines in conversation even though their résumé undersells them.

If a tool cannot appropriately evaluate that person, or explain why it flagged or filtered them, it is not ready for a staffing floor. It is only ready for a demo.

2. It can’t quietly flatten bilingual talent into “a checkbox.”

Bilingual French/English hiring is an important part of the Canadian talent market, and one of the things I’ve consistently found when looking at recruitment technology is that language proficiency is rarely as simple as a résumé field.

Language proficiency isn’t binary. It’s fluency under pressure, in a client meeting, in a second language, in real time. If a platform’s answer to “How do you assess bilingual capability?” is a self-reported dropdown, that’s not an answer.

At BlueSky Personnel Solutions, we actually speak with candidates in both English and French because there is a big difference between saying you are bilingual and demonstrating that you can communicate professionally in both languages.

That gap between what vendors promise in a sales presentation and what their technology actually delivers on the floor is not merely a theoretical issue. In high-stakes recruiting, relying on untested features can directly compromise the candidate experience and derail critical hiring timelines.  That gap between the pitch and the proof isn’t hypothetical.

I recently met with a vendor who told me their platform had 25+ language capabilities. I asked them to show me, on the spot, how it worked in French. They couldn’t. It turned out that particular capability had to be configured on their end first, and they would get back to me.  I am still waiting for that follow-up demonstration today. This delay serves as a constant reminder that glossy feature lists often hide significant operational gaps that only become apparent when you press for immediate proof.

If I can’t ask for a demonstration in the language you claim to support and get one, I don’t actually know what I’m buying. And until someone tests that capability in the real world, neither does the client. When technology is introduced into a hiring pipeline without real-world validation, it creates an uncalculated risk for everyone involved. Until these tools are thoroughly vetted under realistic conditions, recruiters and clients alike are effectively operating in the dark.

3. It needs to tell candidates the truth about what’s happening to their application.

We’ve written before about candidates showing up to interviews coached — sometimes literally fed answers — by AI in real time. That’s a symptom of a bigger trust problem: candidates increasingly don’t know whether a human being will ever actually look at what they submit.

Any technology that makes that black box worse, rather than better, is solving the wrong problem. I want technology that makes the process more transparent to candidates, not technology that makes it easier to filter them out without a person ever seeing their name.

4. It has to earn back time for judgment, not replace it.

The best recruiting tools I’ve adopted over the years didn’t make decisions for us. They gave our team back the hours to make better ones. By automating routine administrative tasks and streamlining baseline data collection, technology should free recruiters to engage in high-value conversations. The true measure of any software lies in whether it creates space for deep human evaluation rather than trying to substitute for human critical thinking.

If a platform’s pitch is “let the AI decide,” that’s a pass. If the pitch is “here’s how we cut your time-to-shortlist so your recruiters can spend more time actually talking to people,” now we’re having a real conversation.

We’ve raised this before: if AI keeps eliminating junior roles, we’re not just changing hiring — we’re cutting off the pipeline for future managers and leaders.  When junior and entry-level positions are systematically removed by automated filtering, organizations inadvertently destroy the foundational learning ground where key industry skills and leadership qualities are forged over time.  Any vendor whose efficiency story depends on removing those first rungs needs a real answer for what replaces them. This is not an argument against HR technology. It’s the opposite.

Recognizing these systemic challenges is not an argument against adopting new HR technology, but rather a call for smarter, more deliberate integration. Modern recruiting tools should enhance human expertise rather than undermine the long-term health of our talent pipelines.

After 25 years of high-touch, merit-based recruitment, I know exactly how much time gets eaten by manual, repetitive work that good technology should be handling.

I’m not walking the HR Tech floor looking for reasons to say no. I’m looking for the vendors who’ve actually sat with these problems instead of simply building around them.

And if that’s you, come find me. I’d love to see what you have to offer.