How AI Helps Companies Match Employers With The Right Employees With Michelle Tinsley
The next time you hire an Uber, you may want to ask your driver what their qualifications are. Chances are they are overqualified. It’s just one of the many indicators of the hard time the world is having with matching employers with the right employees. Michelle Tinsley was so fascinated by her own Uber story that she got the idea to build a platform that would grow into what is now Yellowbird. Yellowbird uses AI technology to match environmental, health and safety professionals in specific localities with companies that are in need of their skills. In this conversation with Chad Burmeister she explains how her company leverages AI in simple yet very useful ways. They also take a look at some parallel applications elsewhere in business. Join in and appreciate just how deeply embedded AI already is in the business world with most people not realizing it.
Listen to the podcast here:
How AI Helps Companies Match Employers With The Right Employees With Michelle Tinsley
I'm here with Michelle Tinsley. Michelle is the Cofounder and COO of a cool company called YellowBird. The website is GoYellowBird.com. The story about the founder talking to an Uber driver and coming up with this idea is cool. We are going to dig into that a little bit and talk about how AIis helping companies match employers with the right employees and that kind of thing.
Michelle, welcome to the show. Thanks for being here.
Thank you. It's great to be here, Chad.
For our audience to get to know you, before we get into AI, I like to personalize the show a little bit and go back to howdid you become who you are? What was your thing when you were a kid? Did you play sports? Where do you live? Share a little bit about your background?
I was basically born in Chicago. I'm 1 of 4kids. I was always the scrappy kid that I felt like I had to fight for a seat. I started kindergarten at age four because my parents were like, “She was getting into trouble at home. We need to keep her occupied and send her to school to go learn something.” I was always trying to keep up with my older brother and older sister. We moved when I was eight to Oregon and finished growing up in Oregon. The nice thing there was tapped into a great school system and I found lots of challenges there. I attended the University of Oregon and then got my first job at Intel. I essentially worked at Intel for 26years, 18 of which were in Finance. That was my primary degree.
My secondary degree was in Marketing. I’ve always got the question of, “You don't seem like a typical finance person. ”They always say, “You are the CFO for this business? I thought you were the chief marketing person.” I'm like, “No, I can do both.” I did move over side ways to be the General Manager of a business unit in our Internet of Things Business at Intel. I enjoyed that role, globally managing a team around the world that provided technical solutions into retail, healthcare casino, gaming and defense. It’s a very broad spectrum but they are taking the standard Intel chips and putting them into very interesting, unique, different use cases.
From there, I became a Product Line Manager and grew a mobile product line in the retail space for Intel. My last assignment was I moved into sales thinking, “If I ever want to run my own company, I’ve got to know how to do sales.” I ran the Internet of Things Sales Team in the United States for six months until there was a corporate reorg. That's when I decided, “Twenty-five and half years, that's a good run. I'm done.” I started my own consulting business. I started serving on the board at a quick trip, which is a major convenience retailer in the United States. I have also been doing Angel investing now for a few years.
Between those three things, I kept myself busy. As I’ve got deeper and deeper in my consulting into the startup ecosystem, I realized, “I would love to be a founder at a company and build something of value. Not just passively invest or give advice but by making it happen on a day-to-day basis,” so I'm fueled by that productivity and utility. I met Michael Zalle in the fall of 2019 and we started YellowBird. We have kept going. We have made it through a pandemic where we can make it through anything then.
Tell us about the Uber story. I remember in an Uber a couple of years ago, I met someone and Bitcoin was becoming a big thing. I think very similarly I said, “What's your deal? What did you do before this?” He goes, “I was mining Bitcoin and then it got to be too expensive so now I'm driving an Uber.” I was like, “This guy was off the charts intelligent.” It was a 45-minute drive across town. He gave me the education that I was trying to get on Bitcoin that I couldn't get and then I realized, “There's probably some overqualified Uber drivers.” That’s how your business started.
AI is the key to profitability. You can always do things the hard, sweaty way but you will need AI if you want to scale.
Our founder got into Uber and the first thing he noticed was the car was spotless. He had chilled water waiting for him. He put his bag in the trunk for him. He was like, “Did I order an Uber? I’ve got like a personal driver. This is amazing.” He told the guy, “You are fantastic. I want to hear your story. What's your background?” That's Michael. He's very open, personable and loves to meet everybody on a personal level. The gentleman told him I'm an ex-Colonel in the Chilean Air Force. I basically moved here to be near my grandkids. I don't want to sit and drink by a pool. I want to contribute and add value. Now, what I can do is Uber.”
Michael let that thought rattle around. He worked with quite a few Environmental Health and Safety folks and said, “These folks entered the industry in the ‘80s when OSHA was created. They are now getting to a point in their career where they want to lightly retire but they still have many skills to offer and so much experience to tap into. It seems a shame.” Quite honestly, when you are a good safety person, you are probably not the best marketer. Do you have all the digital tools at your fingertips to go do that marketing on your own?
His thought was, “What if we created at wo-sided marketplace where we matched together these amazing professional stopped into their skillset, expose them, and then on the flip side, reached out to companies that need better safety but think they can't afford a full-time employee. They can now get exactly the skillset they need when they need it for the job they have at hand. They could get somebody different based on skillset next month or they could keep asking for the same person by name and get them. ”We do cover our workers with insurance. We do want to be a gig done right. We know if we are going to get professional people, we need to do this in a way that honors their experience, background and levels of playing field with our skillset of who we have out there.
There are apps like Fiverr and those that will match me with a $20 an hour to $100 an hour working in another market. Yours is very focused on a very niche market.
We are focused on Environmental Health and Safety work that does include risk management and loss control. I would say about a good 1/3 of our volume now is lost control audits for the insurance industry because folks with a safety and risk background know how to go in and evaluate what a catastrophic risk for a business versus a nice to have is. We do focus on that. We do always use it as a litmus test internally. We just had known now what could have been 200 jobs on the platform but it was no skill. We need a person with a pulse to do something. We said, “That's not our knitting. We need to stick to what we do right and honor the experience that our professionals have.”
Talk to me about where does AI plays a role either now or in the future with the matching piece? I understand that's the big part of where AI may play.
That's what excites us. Obviously, AI, when done right, it's trained on a large set of data. We are still building up our large data pool but we have over 500 match-ready professionals and another 1,500 people coming through our pipeline through our onboarding process. I teach of those people has 10 to 20 years of experience, we've got a lot of data points around the types of skills they have, the types of certifications. There are very detailed, explicit requirements.
For instance, somebody needs to come to teach a fall protection class, but they need to teach the competency level. There has been an OSHA Certification that's required to do that. What AI enables us to do is search through our massive number of professionals, data and pullout the right people, not just for skillset but location. We try to start with doing jobs within a 30-mile radius of where the people live. Obviously, we will cast a wider net if it's a unique skillset. We have had somebody ask for radiation safety. We only have three people on the platform that can do that level, basically consult at a radiation power plant. That's a very unique skillset.
I wonder what some of the other match criteria might look like over time. I think of a friend, Brian Skelton, who does this work. He travels all over tri-state and does this type of training for people. I'm sure his skillset is very niche-y. He's also not looking for work. I would be curious, what about latent people? Is this just for people that are actively looking for work or who does it serve?
We do consider them in inventory. We, for instance, have a woman who only does domestic violence training in the workplace. It's part of the safety curriculum at larger companies. It's a very niche-y that she's the one and only. We are not going to put a big marketing campaign around that. If somebody came to ask us for that or again, she found clients in other locations where she doesn't want to worry about taxes and how to do the compliance, she can put the job through YellowBird.
We do see where in the future AI will help us match beyond the experience and the certifications. We may have five people come up in a 30-mile radius that can do it. What about the personalities? What we have been already asked by customers is, “I want a safety coach, which is very friendly, proactive, let me help guide your employees to a safety culture and then we have had the safety cop.” Depending on the customer and their circumstances, they may want the cop versus the coach.
We want to put personality types or personas like that in there and let the professionals self-select of, “Here's the type of behavioral approach I take to safety.” That will again, let us add more nuance because we look and say, “Maybe 5% of our jobs haven't been a perfect match.” It has been the soft skill is the wrinkle point. Not that they didn't know anything or that they weren't qualified.
There are a few technologies. Crystal Knows can go out and read a LinkedIn profile. There's another one called CodebreakerTechnologies. They go in and they call it the BANK code, Blueprint, Analysis, Nurture, and Knowledge. They stack rank and tell you, “This person is a knowledge first and an action second or blueprint first.” My son is an Engineer. He's going to a school for Engineering. He would be a blueprint first. I'm A for action first and then knowledge second. B is the last thing on my list. Knowing those things in the algorithm is not that difficult to do but very high value.
That's where we have seen also a mix of the types of jobs we do. Everything is short from 1 to 2-hour consult to nine months onsite. It's obviously for the gig staffing where they want somebody more than 2 weeks or 1 month onsite that the personalities become more of a flight. We have said, “Now, the human intervention until the AI is ready is we will let you interview the person via Zoom to make sure that you can ask them the cultural questions and see if they are a fit with the team before we do a formal match.”
I usually ask the question, “What would your product be without AI?” I think until you hit that massively large dataset, you could limp through if AI didn't exist. It could still be a business but where does that tipping point maybe come in?
To me, it's the key to our profitability because I go, “We could do things the sweaty and hard way. We can just throw a bunch of employees at manually matching these things hand-curated but that doesn't scale that.” We do use AI in our platform as a core feature to our engine and basically, get this done in a way initially humans will review there commendations, not an AI. As we train it and get more and more accurate, it will be more automated and need less human intervention.
I talked to someone that takes an image, a snapshot through helicopters, airplanes, drones of power lines all across the US. It turns out something like 76% of all fires are started by power lines. Humans couldn't ever look at all of the power lines. This tool creates a second universe and then builds out AI of, “That's the plant or tree? How fast is it going to grow?” When I asked the question, “What if you didn't have AI?” He was like, “No way.” That's why California is spending $7.6 billion to throw people at the problem when they could spend a heck of a lot less to leverage AI and get the job done. There are a lot of similarities I see across the industry of the kinds of things that you are doing leveraging AI.
Obviously, for my tech background, there’s always been this question asked, “Is tech can replace the people? None of us will have jobs in the future.” I choose the future where tech is enabling us todo things we have never dreamed of. There will always be jobs and needs for humans because we are so creative. I was driving behind an autonomous vehicle and thought, “There are a lot of machinery on that car trying to replicate what goes on naturally in my brain when I'm half caffeinated on the way to work.”
I was like, “To me, it's a marvel of, we will uncover more and more uses for AI but it will then open up more and more possibilities and diverse types of work.” In this case, what we are doing isex posing a very unique, amazing skillset that professionals have to have businesses and have a level of safety they never thought they could afford. Therefore, their employees will be much safer, which means everybody goes home with ten fingers and toes. That's great.
I talked to a company that's called CommSafe.ai. They monitor all of the communications between employees on Slack, Outlook, Gmail and everything, and then they flag things that are extreme or sexual harassment. He said it was like a $500 billion problem in the world. The debate is that big brother or not? You have people that are monitoring it now. What's the difference between turning the lights on? The big question is around ethics. As long as it's monitored in a way that's legal and moral, then, by all means, let's turn the lights on.
AI will eventually be an ingredient like salt. It’s going to be a foundational building block that helps the businesses of the future scale.
As a Manager, I would say, “Can I go have a very early conversation with an employee? Did you realize?” Bring awareness to them because my guess is half the time, they don't even know but they thought it was an innocent joke or ribbing of a fellow employee is borderline harassment. I think once you bring that up, then that makes them aware and it doesn't keep going further down the path to where you get to a place where you have to fight them.
That was another good use of AI. I have talked to almost a hundred founders and CEOs and the level of AI deployment. Most people don't realize that it's going on. You talk to your Alexa or your Siri and you get your chips delivered the same day. I didn't know it was being deployed in my technology like it is at YellowBird.
We are a little fearful of that because when we go talk to investors or potential customers, we will be like, “They mentioned AI. They are to be cool or try to be with the fast crowd.” I think it is an ingredient like salts, where eventually it's going to be a foundational building block that again helps the businesses of the future scale.
What about the sales motion? I'm assuming you are selling to a bigger market but it's at least a targeted audience.
We do have a very targeted persona that the market for EHS services and staffing in the United States is over $50 billion. It is quite a large market but if you have typically been in an office environment, you probably aren't aware of how big it is. We want to get in with decision-makers, typically leaders and operations HR that are making the call on, “Do we do this internally? Do we staff it with a generalist, do we go seek out a specialist or do use a gig economy?” Now, we are trying to get into the evoked set or get into consideration because they don't know this model exists. That's where we have been working a lot on awareness building.
Usually, once somebody puts their first job through, I would say anywhere from 65% to 85%, depending on the month and where the activities at, our customers are coming back for a 2nd or 3rd Gig within 90days. It's because once they know this model exists, they are like, “We haven't updated our distracted driving training in five years. There are a lot of stuff we could put in here now.” Asking us to update policies, procedures, stuff that honestly, employees aren't that excited to do. We can get it done very quickly.
I had a full-time dedicated ops person and I was only leveraging 20% to 30% of the time. I went to him and said, “What else do you think would be worthwhile to do?” I didn't have anything for him todo. We couldn't figure out the other 70%. It made sense to say, “Why don't we move to part-time? You can now work for us part-time and another company part-time, as many as you want.” That's what he ended up doing and it worked out perfectly well. It's almost EHS AAS, As A Service. Where do you think it goes a few years from now, whether in EHS or outside of EHS, talking artificial intelligence in general, think back to your several years in the industry?
I would love to see us be ubiquitous across the US and into other countries. This is a model that's become aware enough that other people start asking, “Why aren't you an X, Y, Z country?” We would have gone international within 3 to 5 years, and then also looking at, are there adjacencies to the work we are doing? Obviously, they lose control. Do we get into ESG overall risk assessments for companies looking at where enterprise risk is coming from? It may not be the employee safety and environmental piece.
It may be food safety. We do have people with food safety certifications now. We just haven't marketed that. I look at it and go, “There are a lot more of these skills that we can expose. ”Hopefully, raise the level of awareness and competency across industries on some of these more serious issues. Food safety is something that could wreck a brand. It's something that could cause hundreds, if not thousands of people to be sick. If we can start eliminating those incidences, that would be great.
It’s an amazing conversation. Congratulations on joining the business in 2019 and leveraging the skills that you have been using over the last several years. Michelle Tinsley, Cofounder, COO of YellowBird. It's GoYellowBird.com. Thank you for sharing with our audience.
Thank you so much, Chad.
Everybody, thanks for joining the show. We will catch you on the next one. Cheers.