Much of the market now files automation and AI under one word. They're different tools for different jobs: automation for the predictable work, AI for the judgment work. Confuse them and you buy the wrong system.
I recently saw a post from Chermaine Dufosse that I keep coming back to.
A lot of the tools now being described as “agents” are really rules-based automation.
And that's not a criticism (rules-based automation is objectively useful, and a surprising amount of hiring depends on it), but her point was simply that we should be honest about what these tools are, because the language shapes how we think about the problem.
.png)
I'd even take that one step further.
The distinction matters because hiring isn't one kind of work. Some parts of it are highly predictable. Some parts require judgment. Those aren't the same thing, and they benefit from different kinds of technology.
The trouble is that much of the market now talks about both under the banner of “AI”. And automation either gets folded into that category, or disappears from the conversation altogether.
Once that happens, it becomes harder to think clearly about what a system is supposed to do. And in my opinion, that's where a lot of buying decisions start going wrong.
If you walk through a recruitment process from beginning to end, you'll notice something fairly quickly.
Some tasks are predictable. We all know that every candidate should get a reply, every interview should be scheduled and confirmed, every reference request should go out when it's supposed to, every SLA should be monitored, and the right person should be nudged before it slips.
The right answer is already known. The challenge isn't deciding what to do, it's just making sure it happens consistently.
However, other tasks look completely different.
Reading a resume and understanding what it actually says, distilling an hour-long interview into the few things that genuinely matter, comparing two strong candidates whose strengths are nothing alike, writing a message that feels thoughtful rather than templated.
Most hiring teams deal with both kinds of work every day, often within the same process. Which is why I think it's useful to separate them. Because once you do, it becomes much easier to see where automation helps and where AI helps.
Rules-based automation is useful because it's predictable: it runs on trigger, condition and action. Once you've set those up, it will do the same thing every single time.
For example, a candidate reaches a stage, so the email goes out. An interview is booked, so the confirmation and reminder are sent. A reference request is due, so the task is created or the nudge happens.
And that's exactly what you want for the routine parts of hiring. You don't want a system to get creative about whether a candidate should receive a reply. You want it to happen reliably, without depending on whether someone remembered, had time, or followed the process perfectly.
AI helps in a different place. It becomes useful when the input is messy, varied, or incomplete. When a resumé is in an unusual format, or you have a shortlist with no clear stand-out, or even when you have multiple candidates that need replies but you don't want them to all sound the same and generic.
That kind of work benefits from something that can interpret context and produce a useful first pass. That doesn't mean it will make the decision for you, but it helps the person making it see the material more clearly.
Simply put, use automation where the work needs to happen the same way every time, and use AI where the work depends on interpretation.
Chermaine describes this as two different value levers: intelligence and execution.
I do think that's a useful distinction, but where I'd push the idea a little further is that most hiring teams don't have the luxury of choosing one or the other. They need both.
Hiring teams need the routine work to happen reliably without consuming time and attention. They also need support when they're making decisions that depend on context and judgment.
Execution and intelligence are both important, but it's even more important to consider whether each part of the process is being handled by the kind of technology that's good at it.
There's another reason the predictable work matters, and it's easy to miss because it's so unglamorous.
When rules-based automation handles the routine work, it also handles it in the system. The reply, the schedule, the reminder, the status change, all of it lands in the record as it happens, consistently, without depending on someone to log it after the fact.
And that matters, a lot.
A clean, comprehensive picture of hiring isn't created at the end of the process by asking people to tidy up the record, it's created as the work happens. The more routine activity takes place inside the platform, the less the story of a hire leaks into inboxes, spreadsheets, and side conversations.
And thatrecord is exactly what AI needs if it's going to be useful.
The interview summary is always going to be better when the system knows the role, the scorecard, the previous feedback, and the stage of the process. The candidate message is always going to be more personal when it understands what has already happened. The shortlist recommendation is going to be strongerwhen it can see the full context rather than a few structured fields.
So you can't just think of a automation and AI as sitting next to each other. The predictable work creates part of the record that the judgment layer depends on.
This is why at Pinpoint, yes we've invested in AI, but we've also invested heavily in the rules-based automation layer, the triggers and conditions that handle the predictable work. And that's what I believe makes the difference.
So I'd probably retire the question most buyers walk into the market with, a variation of “which platform has the best AI?”.
Instead, a more useful question is: what work do you want people spending time on next year? And on top of this, what work do you want the system handling automatically?.
Once you answer that, most of the technology decisions become easier.
Tom is the CEO at Pinpoint, he's passionate about building world-class teams and world-class products for organizations around the world.
