The founders of an AI copilot for recruiting, with a distinctly human approach to hiring.
How do we eliminate bad hiring decisions? In 2018, two entrepreneurs, Siadhal and Shahriah, set out to solve this problem.
If you’re short on time, read the 30 second version of this post.
How do we eliminate bad hiring decisions? In 2018, two entrepreneurs, Siadhal and Shahriah, set out to solve this problem.
Frankly, we'll probably never fully eliminate them because humans and companies are too nuanced never to have bad hiring decisions. We could probably get closer to that than where we are right now, though.
Siadhal Magos, Co-founder & CEO at Metaview
They founded Metaview, an AI copilot, to make hiring more efficient and fair. Despite this, their approach to hiring at Metaview remains human-centric, artisanal, even (as Siadhal describes it). His take on hiring with AI differs from what you might expect from an AI founder.
I am ultimately accountable for who gets hired. Some people think if the machine says it, it must be right. They need to understand that this is not how these probabilistic models work.
Harrison Chu, director of engineering at Arize.AI, agrees. Here he explains how randomness in picking the next token is introduced simply because the most probable option may not be the best response – it just probably is. It's a great example of not accepting that the machine must be right by building guardrails around it.
As everyone rushes to sprinkle AI on every problem, it's good to remember that human judgment and empathy cannot (should not) be easily replaced with technology. In Metaview's case, they put humans at the center of the decision-making process.
Metaview acts as an AI scribe for interviews, providing notes and insights to help (not replace) human judgment in hiring managers' and recruiters' decision-making process.
For Siadhal, accountability in hiring decisions means various things. He wants to vet candidates rigorously. He has strong opinions on what to look for from applicants and how to assess soft skills. He cares deeply about building a world-class team.
In our interview, we discuss:
Metaview's origin story
Metaview's approach to hiring
Founder tips for recruiting
The future of recruiting
If you’re short on time, check out Siadhal’s Practical Tips ✨ below.
1. Metaview’s Origin Story
The founding story behind Metaview starts with a "how might we" exercise. Siadhal and Shahriah used the Google Design Sprint method to align on what problem they wanted to focus on. Given their previous jobs at Uber and Palantir — 2 companies renowned for their rigorous interview processes, it seems fitting that they landed on "how to minimize bad hiring decisions." Neither were recruiters, but they'd both been hiring managers at their respective companies.
I was at Uber. He was at Palantir. This sort of product thinking led us to: what happens if you think of the hiring process as a product? How would you improve it? I don't know because I don't have the data. That made us realize that if we could harness the data from conversations, we could probably move the needle.
They went further and spent a lot of time thinking about 1) why now and 2) why them.
Why now: At the time, more interviews were being held at an arm's length of a camera and microphone. This was the key: better data.
Why them: Uber and Palantir both had rigorous interview processes, but they prioritized different things. Uber was about getting high performers to come in and execute quickly. Palantir prioritized talent density above all else and preferred not to hire unless the candidate was a "bar raiser." Both approaches required rigor, applied in different ways. These perspectives gave them an edge.
Siadhal credits their focus on these two questions with getting them through the difficult times of starting a company.
The thing that kept us going was, fast forward n years – and hopefully, n was a small enough number for us to still exist – do we think people will be capturing the data from their hiring? Whenever we were low, I would think about that, and the answer was yes. And therefore, we should keep on going. The product changed a bunch, but the core thesis stayed the same.
Practical tips ✨
When starting a company, use the "How might we" exercise to identify problems worth solving. Follow it up with "Why now?" and "Why us?" to validate timing and founder fit.
2. Metaview's approach to hiring
I asked Siadhal about his thoughts on hiring and how he feels about candidates using AI in their applications. He isn't bothered by it as long as the final output reads well.
I've got no problem with it. If you elicit a high-quality prompt for an AI to help you write something, you're likely intelligent and diligent enough to get that out. As you probably know, with most tools, if you put in a simple prompt, you get an output that'd be very unappealing to read to someone like me.
What he looks for
What does he look for outside of a resume and cover letter? Our personal view is that those signals are deteriorating in how much they can help hiring managers pick candidates, but Siadhal has an interesting take:
We look at a resume and think: Has the candidate seen what great looks like or, given their starting point, ended up on a really good trajectory? The cover letter is less important in a world where every company is essentially un-opinionated about what they like and has very vanilla culture statements. However, if they are opinionated about their brand, it filters for people who want to fit into a specific culture. Maybe those cover letters could be more relevant in that world because there's less of an obvious answer. You might need spicier questions in your application form, though.
Red flags
I'm OK with work-life balance, and I think it's important. I've got kids, so I get it. However, some candidates equate work-life balance with an easy job. And if that's the case, the candidate is probably a startup tourist who wants a 9 to 5. Turning a small company into a big company is hard work. So, I am mindful of that in interviews.
This is a refreshingly honest take, as it's a timesaver for everyone. Startups are hard and require a level of commitment that isn't for all. Many candidates prefer a 9-to-5, and that's OK. But it's important to have an honest conversation with them about whether they are suited to startup life. Avoiding these hard conversations can lead to much harder conversations later on.
We also discuss the "relevance of experience." It's a common mistake to hire someone with experience from a larger enterprise and think they'll figure it out and adapt. It rarely works out.
We've hired good people, but they haven't worked out because the distance between what they're used to and where we are is just too large. We made the mistake of thinking that we could find a way and that, with enough flexibility and malleability, it could work out. Unfortunately, it never has.
Dogfooding
Unsurprisingly, the team uses their own product for hiring processes. Here's how Siadhal thinks about the product, which also gives a clue as to how he's thinking about Metaview's future.
We're the AI scribe for recruiting. The idea is much more than just a transcription service. Historically, scribes were not just transcriptionists; they were vital to people's communication and decision-making. That's why we think of ourselves as a scribe. There's much more to it than just the notes, but it's a good place to start.
Practical tips ✨
Add "spicier" questions to your application form to filter for cultural alignment.
Be upfront about the demands of startup life during interviews. Look for candidates who understand and embrace the challenges
3. Founder Tips for Recruiting
Here are three founder tips Siadhal shared during our conversation:
Being discerning: It can be tempting to give the benefit of the doubt during an interview. As an optimist, he did this a lot during Metaview's early days. But he's learnt the hard way to be more binary about the outcome.
In every interview , expect that the best version of the candidate is on display. If I come away from an interaction and don't see evidence of X or Y, I take that at face value. Not doing so obviously wastes everyone's time and can lead to hiring the wrong person. People turn up to show the best version of themselves. So when they show you, believe them, and if it's not what you're looking for, it's best to move on.
There's a caveat here. There are many emotional aspects involved in an interview. A candidate can be highly prepared but nervous because of pressure or things not being right at home. In an interview, we look through a small window into that person, which gives minimal context.
A personal example comes to mind of a really strong hire I made at a previous startup, Prolific. The candidate was extremely nervous during an assignment presentation, and her voice was trembling. She was clearly not at ease. I stopped the presentation midway and told her we were not assessing presentation skills; we wanted an insight into her thinking. From that moment on, she read her notes in a conversational manner, and her brilliance shone through.
Expecting the best version of a candidate to show up is more than just a hiring tactic—it's about taking a clear and uncompromising position on company culture. As Siadhal pointed out, being deliberate in your approach ensures that you're not just filling roles but building the right team for your startup. By having a solid vision of who fits, you avoid taking a 'vanilla' approach, that can dilute culture and slow decision-making. It's about making quicker, more confident hiring choices and knowing those who align will thrive with your culture.
Increasing precision: A hiring decision will never be perfect, but that shouldn't be the aim. The aim is to make it more precise and fairer, one that comes from a better understanding of the candidate, by getting more inputs. Looking at the process through this lens, it's clear why Siadhal is building Metaview.
What people do right now is pattern matching based on personal experience, which is also prone to error. In the future, it will be based on patterns across hundreds of thousands of conversations and candidate experiences.
Recommended read: Kochland. A mysterious private company with more revenue than Goldman Sachs, Facebook and US Steel combined. The billionaire CEO has kept the company operating in secrecy with a view toward very long-term profits.
It's an insanely successful company with offices in the middle of nowhere. They manage to attract great talent to a very secluded town. There's something about creating that siege mentality that I admire, even if it's not necessarily the route we're taking. It makes you think about the real clarity of leadership and direction and how the work gets done in a company that can attract the right people.
Practical tips ✨
After each interview, make a clear "yes" or "no" decision. Trust your judgment of the candidate's best self.
Use tools that provide more data points on candidates, such as AI-assisted interview notes. Use insights to make more informed and precise hiring decisions.
Add Kochland to your reading list, if you want to learn about clarity of leadership and direction.
4. The future of recruiting
We discuss the changes he sees in the future of recruiting and how he sees recruiters' jobs evolving.
He talks about a few themes:
Precision
Leverage
Recruiters
On precision:
How can you get to – even if it's never going to be perfect – more precise hiring decisions? AI will enable that through a better understanding of the candidate. It’s also important to clarify what you're looking for in a candidate in an iterative way. Right now, hiring works this way: A hiring manager (HM) outlines what they're looking for, and a recruiter gets back to them with candidates after a defined period. The HM needs to hire but doesn’t necessarily know what they want upfront. They decide based on the pressure they have of needing to hire someone. I hope this is what goes away and becomes more of an iterative process. More precision is required here as the HM clarifies who the best candidate is.
On leverage:
You're going to increasingly have high-leverage people whose job is to ensure you get the best talent because all of the low-leverage parts of that job are automated. And the title of those people will be recruiters. The job can change massively, but the core responsibility is getting great people into your organization and attracting them to it. Assessing that and creating an environment to learn about them and for them to learn about you will still be expensive, even if you do it at scale.
On recruiters:
The variance between a good and a bad recruiter is pretty stark, but I think it will get even larger. Recruiting is a very human activity. We will always want human beings in the company. And I'll always want those human beings to be the best fit possible. To win those people – even if you develop technology or AI that can do 99% of the actual work of finding and assessing – I'll still want to throw a human in the loop to give myself that 1% advantage over the next company. I also want my human in the loop to be better than the next startup's human in the loop.
This point gets to the core of Hiring Humans, and this take is one we fully agree with. The future lies in keeping humans in the loop rather than eliminating them.
There’s a lot of discourse about AI agents replacing recruiters, but we're skeptical. They ‘d need guardrails to keep them on track as they perform chains of complex tasks. As the tasks combine, it’s less likely to achieve the desired output. Agents in more constrained environments can work, but for something that requires the level of context and empathy that recruiting does? We don't see that for a long time.
Conclusion
As we navigate hiring with AI, we predict the most successful companies will strike the right balance between using AI's capabilities and preserving the human elements of the hiring process. Siadhal’s take:
The healthy way to think about using AI is that we're a team, but I'm accountable. We will work on this together to try to get to an outcome. It can feel frustrating if the AI doesn't get to the answer right away. But I'll prompt and adjust the outputs to get us to the answer. Having that approach inherently means that you know that, given you're working on it together, you're the boss and ultimately in control.
Metaview's approach is to use AI to enhance human decision-making not replace it. Keeping humans in the loop and maintaining a relationship-driven approach to hiring shows how AI can make recruiting more efficient and reliable without losing the human touch.
Building great teams is still about people connecting with people—AI is simply a supporting act. Founders and hiring managers will have to embrace the precision and efficiency that AI tools like Metaview offer, while maintaining the human skills and judgment that make great hires possible. This will be a balancing act, but one that promises to revolutionize how we build teams and companies.