Transforming Hiring with AI

Transcript

Thank you everyone for joining us for this chat. My name is Prem Kumar, CEO and co-founder at Humanly. I’m joined by Anthony Roli, CEO of Interview AI. Thanks for joining us, Anthony.
Excited to be here with you today, P.

Awesome. Well, I know we’ve had a lot of discussions about many of the topics related to recruiting, AI, interviewing, screening, and scheduling of job candidates. One of the things I want to dive into is candidate experience. Unfortunately, for high-volume roles, there are many unfortunate candidate experiences like not hearing back, not knowing next steps, and being ghosted. Sometimes this applies on the other side as well. How do you think about candidate experience, either from your past TA roles or in your current role? How do tools and technology help or hurt?

You know, it’s interesting. I think all good talent people have a commitment to candidate experience. It’s often something we want to get right. But I found that, in both recruiting and leading recruiting teams, it’s one of those things we want to get right, but there’s a lot to manage in recruiting. You want to make sure you have a great process, that you’re rigorous in your analysis of candidates, and there’s just a lot going on. This often means that candidate experience falls through the cracks despite good intentions.

It’s fascinating to watch the rise of AI and how it’s impacting recruitment. AI gives humans more scale than they’ve ever had before. It allows us to scale our touch, making candidate experience better by setting expectations and being timely with communication. Sometimes we overcomplicate candidate experience, but it boils down to proper expectation setting and timely communication. Treat candidates kindly and care for their needs. The technology now allows us to do that at a greater scale, touching more people with fewer cracks in the process.

I think that’s a great point about scaling touch. Often, just a few touches and proper expectations are enough. Candidates just want to know, even if they didn’t get the job or aren’t moving forward, that their time was respected. Tech can make that possible. If hiring teams had unlimited time and resources, they’d speak with every candidate that applies. Technology can help make that feasible. Another important thing is building long-term relationships with candidates, even if they aren’t right for the current role. How do you normally go about measuring candidate experience? Surveys? What’s worked for you?

Candidate surveys can be a really important part of the process for qualitative feedback. But I also think there are quantitative measures as well. Knowing your funnel metrics—like how long people are waiting for a response—is vital. A platform I encountered once showed the average time candidates waited for a response, and it was often much longer than people expected. That’s something to keep track of. Additionally, creating internal candidate commitments and publicizing those to set expectations can help, too.

Absolutely, and pairing that quantitative data with behavioral data and qualitative insights is essential. Some companies, like Virgin Media, quantified the loss of candidates due to poor experiences—up to $6 million per year because candidates who had a bad experience stopped being customers. Particularly in B2C, your candidates are often also your customers, so there’s a real reason to get it right.

The next topic I want to explore is decision-making in HR and recruiting tech, especially with AI. How do you think tools can help make us better interviewers, and how can they assist in making better hiring decisions using technology and data?

I’m really interested in decision-making in HR and how data and AI help with that. When it comes to hiring, I believe the more standardization we have, the better. Interview AI, for instance, started with better questions, but the more important part is helping interviewers assess the answers. Standardizing processes and making sure all candidates receive a similar experience are key. Having consistency in your methods and rubrics for assessing interviews makes a big difference. Often, hiring managers are so eager to fill a role that they rush the process, so having structure in place ensures the best decisions are made.

What intrigues me about tech is its ability to capture more data than humans can during interviews. AI can help capture signals we might miss. It’s tough for one person to judge another’s abilities in an artificial interview environment, so AI can help us gather more insights. I know you’ve been interested in that space, too.

Yes, especially with remote interviews over Zoom and Teams, more can be measured, and what you measure can influence improvement. For example, we analyzed several hundred interviews and found that candidates who were in the same city as the interviewer received higher overall scores and had more small talk, which likely influenced the final decision.

I want to ask you a tough question about intuition in hiring. How does intuition play a role in making the right decision, and how does it mix with objective data and AI?

It’s something I’ve spent a lot of time pondering. Some hiring managers rely heavily on intuition, which can be fraught with bias. But we owe it to ourselves to assess data better. AI can help bring more data points into the mix, and humans can still play a role in measuring those less tangible elements, like cultural alignment. We need humans in the room to measure what isn’t on paper, but we can guide intuition better by asking more questions. Raw, unfiltered intuition isn’t helpful, but if we dig into why someone feels a certain way and examine those reasons, we can use that information meaningfully.

I liked what you said about guiding intuition. Just like with ethical AI, where we demand transparency, we should demand the same of ourselves when it comes to intuition. If we can be transparent about why we feel a certain way, we’ll learn a lot about ourselves and how we make decisions.

Switching gears to tools now, I know you experiment with different tools in both work and personal life. What do you look for when evaluating tools that use generative AI or newer technology? How do you assess their impact in the TA and hiring space?

When evaluating AI tools in HR tech, I first look at how responsibly they’re implementing AI. How much decision-making authority are they giving to the AI, and how crucial are those decisions? For example, in Interview AI, we propose questions and suggest what to look for in answers, but we’re cautious about saying, “Here’s what you should do based on this conversation.” We’re not there yet. I want to see companies being cautious about not removing humans from crucial decision points.

Once I’m satisfied with that, I dig deeper into how the technology works. Is it using off-the-shelf models or proprietary ones? How is data being protected? These are important factors when evaluating vendors.

That’s a great point, and I think when it comes to using AI in hiring, it’s not about removing humans from the process but augmenting the human role to make better decisions. I also think we need to approach AI adoption in the same way we would with hiring people. You don’t hire someone without a clear role or purpose, and you shouldn’t bring in AI without knowing exactly what problem it’s solving.

Exactly. Writing a job description for your AI can be helpful. It clarifies the problem it’s meant to solve and how it fits into your process.

Any misconceptions you’ve seen around AI, whether for better or worse?

The biggest one is that AI is going to take all our jobs. The more I use AI, the more I realize it’s a tool that makes us more efficient, but we still need humans to interpret and guide it. AI won’t replace us; it will empower those who adopt it early. I also think there’s fear around bias in AI, which is legitimate. But we need to compare that to the bias humans already bring to the process. I’m curious to see how AI will help us overcome some of our own limitations.
I agree. We don’t scrutinize or audit human interviewers the way we can with technology. That’s one of the benefits of AI—being able to audit it more deeply without creating cultural overtones.
Absolutely. I appreciate this conversation, Prem. Looking forward to more of these in the future.

Same here. Thanks for your time, Anthony!

‍