12.06.2023: Demo Day -Sonar Talent Intelligence

CHRIS MANNION: Hey, everyone. Very excited to be here today. My name is Chris Mannion. I was a MBA class of 2016, and the founder of Sonar Talent. Now, we're an AI based talent operating system. Our goal is to improve employee retention and performance by utilizing current employees to fill critical skill gaps on the team.

I wasn't always in the kind of recruiting and HR space. I moved across about five years ago. But my background is in supply chain analytics and aerospace engineering. So I'm actually pretty excited to be working in this new space and trying to fix some of the problems that I saw firsthand in my last role.

So more than ever this year, department heads have been struggling to meet pretty ambitious goals with very constrained budgets. External hiring actually used to be the primary go to if you needed to fill a skill gap on your team.

But because competition is so fierce, with now two open roles for every job seeker in the market, the challenge of actually filling these roles with external hires has gotten more and more difficult. So the skill gap is only going to get worse over time.

We actually expect to have about 85 million open roles or skill gaps in the market by 2030. And especially in key roles where machine learning, AI, and engineering is involved, you're going to see this skill gap get even more severe.

Now, the cost to a typical business, we looked at some Fortune 500 companies, is pretty substantial. If you go all the way back to 2009, when we had a huge kind of a post recession hiring spree, you could actually hire someone in about two and a half weeks. Now, as the competition has increased, that has gotten much more difficult.

And so the cost of an open role right now has actually more than tripled to about $160,000 for each open role. Now, if you look at any of the big companies in the Fortune 500, that's a $100 million plus gap on their balance sheet for lost revenue and lost productivity.

And so this also doesn't account for the increased cost to employee attrition based on the gap roles actually causing additional workload for the people that are currently in the role. And we've seen that increase over the last couple of years with the great resignation.

So right now, business leaders are trying to solve that problem through one of three ways. The first way is to actually try and build their own analytics capability in house. So using spreadsheets, they can try and identify where they have skill gaps on their team, and who's potentially able to fill those gaps, and actually move them around to try and plug those holes.

But this takes those leaders and those management teams away from their core jobs. And quite often, they don't have access to the data that they really need in order to drive that performance.

Next, they could actually use one of the existing HR technology tools that's available. These generally are built on historic internal data. So they can give good reporting capabilities, but they're pretty expensive, take a long time to implement, and generally are not very prescriptive and helpful when companies are expanding into new areas.

And then, finally, they could ask the HR team to try and solve this problem, to maybe build out a bigger analytics capability. But HR teams are already overworked, and their budgets are so constrained that headcount is very unlikely to be allocated without a critical need.

And so recognizing this problem, we've built a machine learning powered solution that lets department leaders and their management teams quickly identify skill gaps using predictive talent intelligence. So what they can do right now is analyze all current employees to find internal candidates with the right skill set, experience, and competencies to excel in a role made by a skill gap on their team.

Users can manage the internal recruitment process either through our app, which is fully online, or directly through the applicant tracking system, if they currently have one in place.

And then, finally, one of the more recent implementations is a generative AI based Slack app, where we actually allow current employees and managers to communicate directly with the process using their current workflow, without having to go to any external systems.

So the product actually comes from a lot of experience that I personally had in my last role. So one of the final projects I did was actually building out an internal mobility process at Wayfair. Wayfair is a Fortune 500 company. At the time, we had about 20,000 employees globally.

But the CHRO had an issue. They had these highly critical roles across the teams, but a budget constraint and a headcount freeze that prevented us from hiring externally. So what we had to do is figure out how do we actually solve those skill gaps without increasing headcount. We actually were trying to decrease headcount over time.

So what I did was actually help to coordinate a whole internal process change, and then direct a lot of the internal high potential candidates into those roles to both decrease attrition of those top performers, but also fill the gaps very quickly.

We were actually able to fill the gaps within 30 days of launching this initiative. We also reduced attrition within that top performing cohort. And then those people that actually moved into those roles actually had stronger post higher performance, because they were internal candidates and were already well versed in the types of work that needed to be done.

So what we're seeking right now is the next stage of our pilot program. We spent about two years perfecting the tools and the models, working with early stage startups where we've been helping them to identify external candidates. And as we're moving to apply this same process internally, looking for three teams in large organizations that have critical skill gaps that need to be addressed very quickly.

The algorithms that we've built have been based around business product and engineering roles. But we think that we can apply them very easily to any other types of roles. We just need a little bit more onboarding time. And we'd like to work with a customer who would be willing to invest in making that happen.

The pilot we're going to run for three months. And the goal is to show how we can increase post high performance and reduce attrition of top performing cohorts with this process.

So that was the end. Happy to take a couple of questions, and I have some additional slides as well to go through in the breakout room, as well as a product demo.

ARIADNA RODENSTEIN: Fantastic, Chris. Thank you. Can you say how this is different a little bit more from like a GPT only solution? Like ChatGPT.

CHRIS MANNION: Yeah. Great. So ChatGPT is trained on large language model, which is essentially most of the publicly available information that's out there, which is actually a great way to build out natural language processing capabilities.

And actually, a lot of existing HR systems have started to incorporate GPT models into their platforms. And then the big difference is we've built everything from the ground up to be focused on the talent space, to really focus on best in class processes to improve post higher performance for employees.

But I think I'd say the big difference is that we are focused on the predictive capability. So we're not trying to say, you know, this is what a job description looked like in the past, or these are the types of resumes that you've received in the past.

We actually want to start with what is the kind of skill set that is most likely to result in top performance, and how do we find those people. So we've built models from the ground up. And we actually do corporate generative AI capabilities to replace a lot of the early NLP models that we built out. So we are using kind of both hand in hand.

ARIADNA RODENSTEIN: OK. And how would you say-- what is an ideal candidate? What do you focus on?

CHRIS MANNION: Yeah, so it really depends on the type of company. There's kind of like one way of doing it, where you just kind of look generically at who went to the best schools, who went to the best companies. That actually really narrows the available talent pool.

What we actually do is look for analogous companies that have grown in the past. And we look at the cohorts of people they've hired to identify top performers within those cohorts.

What we do as a shortcut right now is look at time between promotion. So we look over time series and figure out, how quickly are people being promoted. Then we actually drill down into the traits that those people have in common.

And we also look at the bottom quartile performance as well. And I'd be happy to go into more detail in the breakout room as well. It can get pretty in-depth, and I'm very excited about it as well.