Most of what I read about AI in recruiting falls into one of two buckets: hype (“AI will replace your recruiting team by 2027, click here to find out how”) or fear (“AI is coming for your job”). Neither one really matches what’s actually happening inside teams that are doing this well.
So instead of writing just another “hot take,” I sat down with someone who’s actually living it. Rethna is a Director of Talent with 16 years in recruiting, and she’s spent that time building talent functions from seed stage to Series F, with over 3,000 placements across sales and senior-level hiring along the way.
I wanted to know how someone with that much runway actually thinks about AI. Not the marketing version. The day-to-day, in-the-trenches version.
Normally I’d take a conversation like this and turn it into a story, folding in my own take and what I’ve learned along the way. But Rethna’s answers were too sharp, too specific, and too honest to filter through anyone else’s voice. So here it is, mostly in her own words, organized around the biggest theme I walked away with: AI should earn its place, but there are parts of recruiting where the line shouldn’t move.
Where the Line Never Moves
This is the part of our conversation that sparked the headline, and I think it’s the most important thing a talent leader can say about AI right now.
Q: What work do you believe should always remain human?
“We have a strong bias that interviewing should always remain human. We will never replace a first screen with AI. We will use AI for notes generation and review, and we will always have a human review of anything that an AI generates.”
No hedging, no “it depends.” That’s a policy, and it comes from someone who has automated plenty around it. I personally agree with that. Like I always say, call me “old school,” but I don’t automate first impressions.
Additionally, I see plenty of companies use AI as a first screen but never circle back to actually review what came out of it or how to use that information. For example, a friend of mine took an AI interview for a sales manager role, only to learn later the same transcript was also used to evaluate her for an operations opening, one she was just as qualified for, but hadn’t tailored her answers toward. That’s a mistake in my opinion and it’s nice to see a team that has this down pat.
Q: Where have your expectations for AI been challenged?
“We are still experimenting, and sometimes we have sensitive discussions with our hiring managers, and the discernment of everything being transcribed and reviewed is pretty important. We have a human-in-the-loop workflow, so everything is carefully reviewed before being sent anywhere, and we are very mindful of the audience. This is an area where we had to live and learn. We have encouraged folks to experiment, which is key in general to us learning.”
That’s the balance a lot of “AI in recruiting” content misses: experimentation and discipline, held at the same time. It’s giving the team room to try things, inside a review process that catches what shouldn’t go out unchecked.
Where AI Earns Its Keep
With that line firmly drawn, here’s where she’s actually let AI in, and how far.
Q: How are you currently using AI across your talent acquisition organization?
“We are using AI and have been using it consistently for over a year. We are proud to say that we are using AI to drive down review time, capture candidate notes, and make sure our candidates are having a good experience. We are using AI for sourcing, and making sure our JDs are consistent, as well as being able to drive the most value for recruiter time. We are having sourcing agents embedded into our sourcing, since over 90 percent of our positions are sourced, simply because of the niche nature of the roles we recruit for.”
I personally can get down with that. I like to leverage this technology to remove manual efforts so I can spend more time on the phones.
Q: Which AI use cases have delivered the most measurable value?
“Our intake agent joins intake calls to take down the administrative need. We have a history of how many times we opened the role, and the agent comes and brings in all the notes to the intake meeting, so we have all the data points we need to seamlessly open a new requisition. In a small team, organization is paramount to make sure we have everything outlined in one place.”
That’s the front end handled. But the same discipline shows up on the back end too, in the follow-through most teams let slip.
Q: What parts of a recruiter’s day do you believe AI should own?
“We do think AI tools would be able to make the footprint of a recruiter much more expansive and well-organized. So much of being a good recruiter is good follow-up.”
And honestly, I could not agree more.
The Operations Layer
This is the part people don’t talk about enough, not the flashy sourcing-agent stuff, but the structural work AI is doing underneath a lean team.
Q: Has AI changed the way you think about recruiting metrics, forecasting, or capacity planning?
“AI has made the admin around these things (clearer insights into recruiting metrics and forecasting, capacity planning) a lot more dynamic and accessible. What used to take hours in terms of going through the data now takes a couple of minutes.”
Q: Where do you see the biggest opportunity for automation in TA operations?
“Regular metrics reporting, and getting insights and trends with that sort of reporting, having themes that follow with job openings. How many times are we filling roles again? What can we do to add more efficiency to the process? etc”
The Takeaway
This conversation covered a lot of ground, from the sourcing and admin work AI takes off her team’s plate, to the metrics and forecasting that used to eat hours, to the guardrails she keeps around anything candidate-facing. Every use case ties back to giving recruiters more time for the part of the job that’s actually irreplaceable: judgment, follow-up, and the overall candidate experience.
A huge nod to Rethna for shedding some light on her approach.