Skip to content
April 4, 2023 · 15 min read

Season 3, Ep. 19 – AI that can tell if you’re sober, with Ken Fichtler, CEO of Gaize

Breathalyzers have existed, in one form or another, for almost 100 years and have helped determine safe usage levels of a legal intoxicant. This week Ken Fichtler, CEO of Gaize, talks with Faith about the growth of legalized marijuana, and how its relative difficulty to test for in safety situations led to the creation of the company’s AI testing system.

[email protected]


Read transcript


Faith (00:05):

Hey, Ken. It’s really nice to meet you.

Ken (00:07):

It’s nice to meet you.

Faith (00:08):

Where are you based out of? You’re in Bozeman?

Ken (00:11):

I’m in Missoula.

Faith (00:13):

Missoula? (Ken: Yeah.) Very cool. I haven’t been out west very much, but one of my best friends here in Nashville spends the summer, usually, out in Montana. (Ken: Nice.) So maybe this is my summer to join her <laugh>.

Ken (00:25):

Yeah, definitely. Summers in Montana are the best.

Faith (00:28):

That’s what I hear. I’m from New York, originally, and it’s similar. Definitely buggier, but, (Ken: Yeah.) you know, Nashville’s just, it’s like the inside of someone’s sock in the summertime. It’s like, hard to breathe.

Ken (00:42):


Faith (00:43):

You know, so I definitely, I crave those mild summers, but yeah. It’s really cool to have you. (Ken: Thank you.) I’m stoked. All right, well, we’ve got Ken Fichtler today on the Frontier podcast. Ken, you’re the founder and CEO of Gaize.

Gaize promo video voice actor (00:57):

(UPBEAT MUSIC PLAYS) We believe that every driver on the road and every worker in a safety-critical position should be sober. That’s just common sense, and that’s why we built Gaize, a scientifically proven tool for detecting impairment from cannabis and other substances. It’s a non-invasive, automated, rapid test that measures eye movement using the same tests that have been used by drug recognition expert police officers for decades. (PROMO VIDEO AUDIO ENDS)

Faith (01:24):

But at a high level, Gaize is the first real-time test for impairment from cannabis and other drugs. And on my notes, I see that Ken lives in Bozeman, but I know that Ken is actually in Missoula right now, and previous to this, before Founding Gaize, you were the Chief Business Development Officer in the cabinet of Governor Steve Bullock’s administration. (Ken: Yep.) Which is, I love a career change story, so it’s not on our agenda today, but we might have to talk about that.

Ken (01:56):

Yeah, definitely.

Faith (01:57):

And you founded Gaize in 2021, right?

Ken (02:00):

Correct. Yeah.

Faith (02:00):

Awesome. Well, let’s start with just, you know, what is Gaize? What is it that you’ve built?

Ken (02:06):

Sure. So Gaize is a real-time impairment screening device that’s based on eye movement. So we run a series of eye movement tests, and we analyze those using machine learning and statistical models to understand whether or not the eye movement that we’ve recorded is displaying the signs and symptoms of impairment that we know to look for.

Faith (02:26):

That’s the most…I can tell that you’ve practiced your elevator pitch <laugh>.

Ken (02:32):

<Laugh>. Thanks.

Faith (02:34):

Okay. So, moving from obviously politics into, A, being a founder of anything, but B, specifically, the founder of this, is quite a move. So I’d love to hear, you know, what was the impetus for that move? Kind of what inspired you to build Gaize?

Ken (02:50):

Yeah. I mean, I’d been in private industry for most of my career, and so the first real weird career transition I did was from private sector to government. (Faith: Mmm <affirmative>.) So I was recruited to come be the Director of Economic Development for the State of Montana by Governor Steve Bullock, and that looked like a really interesting opportunity, so I decided that I should do that, and it kinda seemed like a once in a lifetime thing. And it was amazing, and very rewarding, and fun, and challenging, and all that. And so while I was in that job, I really got exposed to the problem that we’ve now solved, which was Montana was a medical cannabis only state at the time. I thought it was probably gonna become a recreational state, which has since come to pass, and so I was trying to understand what was gonna happen if this took place.

Ken (03:37):

And so I was talking with counterparts in states that had already legalized. I was talking with business owners, and law enforcement officials, and legislators in the State of Montana trying to understand what the impacts in the state would be, and it all sounded, you know, pretty good overall. Like, there’s a lot of tax revenue available here. It makes a lot of sense if like, alcohol’s gonna be legal for cannabis to probably also be legal. But what kept rising to the surface was this issue of no device that was available to check for impairment, (Faith: Mm-hmm <affirmative>.) and so that struck me as, you know, both a really important safety challenge, and something that, you know, ultimately would risk the entire legalization effort if things were to go poorly, or go sideways, or there were big safety incidents. And then lastly, it was a huge, you know, economic opportunity, a big business opportunity. So my job was over when our governor was termed out, and so this was, by far, the most interesting problem that I had seen, and thought I would jump in and tackle it.

Faith (04:36):

That says a lot, because I know, obviously, working in government, there’s a lot of problems that probably come across your desk. (Ken: Oh, yeah.) So the fact that this is kind of the most exciting says a lot. It’s been over a decade since the first states legalized recreational cannabis use, and obviously, during that time, there’s been a lot of changes to, I mean, just the number of states that have followed suit, but also the way like, the impairment testing landscape, right? (Ken: Yeah, yeah.) And so I’d love to hear your take, as someone who’s kind of been in it, what have those changes looked like?

Ken (05:12):

Yeah, I mean, so when the first state started legalizing medical cannabis about a decade ago, as you mentioned, it didn’t look like a huge issue. You know, it was confined to a relatively small number of people, relatively small number of states. And so, as legalization has proceeded, it’s really sort of become a bigger and bigger issue where you’ve had, (Faith: Mmm <affirmative>.) you know, recreational access, people are consuming cannabis less and less responsibly, the social stigma is declining, and so the usage rates are increasing. Of course, the potency of cannabis has also increased, and so it’s an incredibly impairing substance, and it’s very, very different than how alcohol works. It’s very challenging to understand how impaired someone is using traditional means. (Faith: Mm-hmm <affirmative>.) And so the tests that we’re using are the same ones that law enforcement, drug recognition expert officers have used for about 45 years.

Ken (06:02):

So those are the “track my finger” tests that you’ve probably seen on TV, (Faith: Mm-hmm <affirmative>.) and what they’re looking for are these very subtle eye movement changes that happen as a result of impairment. So it struck me that these tests were better performed by a robot than a human, (Faith: <Laugh>.) and so I thought, okay, well, if we could do that, is there a way that we could also analyze the data that we could capture? And the answer to that was “yes,” and so that’s really what we’ve built. And so the evolution of detecting cannabis impairment has started from, you know, there’s no device at all, to now, there’s a couple devices that are sort of trying to be impairment detection devices. I think we’ve nailed it. I think the other ones probably haven’t as much, and,  you know, we’re also the first on the market with something that is scientifically validated and appears to be extremely accurate.

Faith (06:58):

So many just like, nodes within what you just said <laugh>. (Ken: <Laugh>. Yeah.) That is fascinating. I mean, first of all, I appreciate your confidence that the only time I’ve seen an impairment test is on TV, and that I’ve never had one myself. (Ken: <Laugh>.) I really appreciate that. It is true, but, you know, I’m glad that that’s the vibe I give off, but…(Ken: <Laugh>.) <laugh>. You know, it’s also fascinating that until, feasibly until Gaize, right, the humans were, we were relying on humans’ ability to really like, process data to make a determination for, I mean, this kind of like, massive safety concern, and that’s just incredible to me. Yeah.

Ken (07:46):

And it’s so hard, yeah. Totally. And there’s, you know, there’s human error, there’s subjectivity, there’s, you know, poor, sort of, testing conditions, and that, you know, there’s just innumerable ways that this can go wrong if a human is doing it. So by eliminating all those, we think, you know, there’s an enormous amount of accuracy to be gained.

Faith (08:01):

And I think it’s also important to note that, you know, safe consumption is at the core of really what you’re doing here, right? (Ken: Yeah.) Like, how can we pursue legalization while maintaining like, public safety and also just safety of the person using (Ken: Exactly.) the substance. And you’re so right. I mean, you know, alcohol is one thing. The most diluted we can get with, or not diluted, it’s the opposite of diluted <laugh>, (Ken: <Laugh>.) the most potent we can get with alcohol is still, you know, it’s easy to measure (Ken: Mmm <affirmative>.) how much you’ve taken, and the same is not true with marijuana products. So it really is, you know, it’s a huge problem space. I’m curious how you see what you’ve built at Gaize translating into other impairment areas, beyond legalized marijuana, as your company grows.

Ken (08:55):

Yeah, it’s an incredibly important point. So the tests that we’re using have been shown to be sensitive to impairment from every class of substance, so any kind of drug you can consume has a unique bio-signature that can be found in eye movement, typically. (Faith: Mmm <affirmative>.) So what we’re doing, we started with cannabis; we conducted the world’s largest clinical trial that’s ever been done on cannabis impairment, and we are now moving from cannabis into other drugs. So we’ve already captured some alcohol data; we’re capturing ketamine data. We’re gonna go on down the line and capture data from additional substances, and over time, the product will become much, much better and able to get down to, we think, very pinpoint accuracy with most substances. (Faith: Mmm <affirmative>.) There are some that, I think, are gonna be very challenging. So, for example, inhalants is a category that is so broad that it’s very difficult to, I think, classify the impairment that we’re gonna see as an inhalant-related impairment. It’ll look like impairment, but we’re not gonna probably be able to say, this is, you know, an air duster or, you know, paint or whatever <laugh>. (Faith: Mm-hmm <affirmative>.) There are some things that are just gonna, ultimately, be confounding, I think, to any test. But the vision, really, for Gaize is to be a single platform that can detect impairment from any class of drug, and so that’s really the world that we’re moving towards.

Faith (10:16):

That’s fascinating. And I imagine for users who are not…technology companies who sell to other technology companies, usually the sale is a little bit easier, because we’re in a constant iterative state, and we kind of move at the same pace, but technologies that are selling into industries that are not, kind of, your typical technology buyer, law enforcement, you know, (Ken: Yeah.) to be blunt here. Like, I think the value of having a single flat platform really cannot be overstated here.

Ken (10:48):

Absolutely. And, you know, law enforcement has a long, sort of, notoriously long and challenging sales cycle, but it’s also an incredibly sticky customer, so, (Faith: Right.) we think that that market has incredible potential. And on the other side, you know, there’s businesses. Any business that has safety sensitive employees is certainly something that we’re interested in tackling, so we’re actually finding our first customers in the commercial space (Faith: Oh, wow.) in industries like construction and manufacturing, oil and gas, you know, the kind of situations where if you’re showing up to work high, you could kill the guy next to you. And cannabis impairment is something that has never been really detectable by businesses. You know, of course, law enforcement have these officers that are especially trained. (Faith: Mmm <affirmative>.) Businesses don’t have anything like that, and so the (Faith: Right.) the response has been, “Well, let’s just ban all THC.” Well, now, THC is a legal substance in many states, and so you can’t simply say, “Well, we’re gonna ban you from using a legal substance.” (Faith: Mm-hmm <affirmative>.) And so the need for detecting active impairment from cannabis is incredibly important and getting more so every day. So our customers in the commercial side are really excited to finally have some way to allow their employees to live their life on the weekend, but also come to work and be verifiably sober.

Faith (12:10):

Yeah, and, you know, you mentioned Gaize conducted the largest, you know, to date, the largest clinical trial (Ken: Yeah.) on cannabis impairment. Not many founders can say that they had that experience. So I’d love to hear anything from that experience that really kind of has stuck with you.

Ken (12:29):

Yeah <laugh>, I mean, that was my first clinical trial, too, and (Faith: <Laugh>.) the science in the space is incredibly nascent, and so we didn’t set out to have the largest clinical trial ever. It just sort of happened that way. We needed a certain amount of data, and that number of participants was the largest to ever have participated in any single trial. (Faith: Wow.) So we had 350 participants. We brought them in sober, theoretically. We actually found some that were not sober when they came in. (Faith: <Laugh>.) We measured their eye movement in the sober state, and then we got them high using recreationally available cannabis, we did this in Canada, (Faith: Wow.) and then we captured eye movement readings four additional times, and so we have a really clear picture of what it looks like to go from sober to impaired and what the impairment cycle sort of looks like.

Faith (13:14):

That’s fascinating. It reminds me, I think if you were to ask my partner what the best day of his life was and not give him any context, (Ken: <Laugh>.) he would say it was when he was living in Denver, I mean, like a decade ago, and the police department paid him to come in for a day and just get super high, (Ken: Yeah.) so they can kind of see the effects of like, different levels of cannabis. And, you know, I mean, that’s, that’s an example of like, what a really profound need there is for something like Gaize. (Ken: Absolutely.) You know, the best they could do is just get my partner super high and be like, “Look, this is a high person,” (Ken: Yeah.) “Watch out for this.” Like, that’s incredible, and what an incredible pace of innovation (Ken: Yeah.) led by you and your team

Ken (13:59):

That’s called a “green lab”, those law enforcement (Faith: <Laugh>.) trials where they get somebody high, and they try to figure out, you know, who’s high, and those are happening all over the country. We’re actually participating in several, (Faith: Wow.) one, including, coming up this Saturday. And so…

Faith (14:13):

Oh my gosh. If you need a participant, call my boyfriend.

Ken (14:16):

Yeah, perfect <laugh>. These are, you know, valuable training tools, but it’s also, I think, really illustrative to law enforcement officers how difficult it is to discover cannabis impairment. Like, frankly, their accuracy is just not that great. (Faith: Mm-hmm <affirmative>.) So it’s a very subtle substance, much, much more so than alcohol, and the kind of very fine grain, highly focused attention to detail you have to have to be accurate is very difficult. (Faith: Right.) So there’s a need for a technology device, whether that’s Gaize or, you know, something else, to solve this problem, I think.

Faith (14:51):

Yeah. So we think a lot about team building, obviously, just by the nature of what we do here at, (Ken: Yeah.) and we talk to dozens of founders, dozens of technology leaders every day who are coming to us to hire developers, and everyone’s got a unique situation, right? And it’s difficult in its own regard, but when I think about the hiring situation at Gaize, you are, it’s a combination of AI, eye movement tracking, scientific research, which we just talked about, (Ken: Right.) and, of course, product development, and all of that’s required at like, an expert level, right, to create something like Gaize. (Ken: Yeah.) So I’m curious about what the process of building your team looked like.

Ken (15:37):

It was, I mean, it’s been challenging. It still is challenging you know, finding, I think, one, the people that can do the work, of course, is a challenge in and of itself, but then I’m also a big believer in passion and interest in the subject matter, and so finding that, then, second level of, “Okay, you can do the work, but do you also care about doing the work?” (Faith: Mmm <affirmative>. Yeah.) has really been difficult. I mean, it took me a long time to hire my CTO, many months, and then hiring our sales team took several months. Hiring our developers has been challenging. It’s, you know, it’s a fairly niche area. There’s not that many developers that have worked on the, sort of, tech stack that we’re using, and we’ve got a web app, we’ve got a mobile app, we’ve got AI, we’ve got the, you know, various connections that have, we’ve got the VR application, we’ve got all the various connections that have to happen between them. So I…

Faith (16:30):

Oh my god. I didn’t even think about VR <laugh>.

Ken (16:31):

Yeah. It’s a reasonably complex product, actually, and (Faith: Yeah.) yeah, so finding the right team members has been challenging and remains that way. But I’m really proud of the team we have. I think we’ve assembled a group that’s done an incredible job on, you know, ultimately, pretty limited resources. So (Faith: Yeah.) very proud of what we’ve built and who we’ve got doing the work.

Faith (16:56):

That’s awesome. And what, you know, what a testament to passion, because you know, in my mind, the largest hurdle here is expertise and technical ability, and the fact that right off the bat you’re like, actually, it’s been, you know, passion is really the strongest indicator for who does well here, (Ken: Yeah. ) and that’s what we screen for. (THE FRONTIER THEME FADES IN) I think that just says a lot about how you build products and build your team, so kudos to you. (Ken: Thanks.) Well, Ken, this has been fascinating. This is very far outside of any industry I’ve worked in before. If folks wanna get in touch with you, learn more about Gaize, where should they go?

Ken (17:36):

Yeah. Our website is Gaize,,, and my email is [email protected]. So we’re available there. Yeah, I would love to connect with anyone.

Faith (17:51):

Perfect. Thanks for listening to the Frontier podcast, powered by We drop two episodes per week, so if you like this episode, be sure to subscribe on your platform of choice, and come hang out with us again next week, and bring all your internet friends. If you have questions or recommendations, just shoot us a Twitter DM @theFrontierPod, and we’ll see you next week. (THE FRONTIER THEME ENDS)