Trey Holterman (CEO and Co-founder, Tennr)

Can you explain your job to a five-year-old? 

I've got this big stack of papers, and on it is a bunch of information that I really need you to read. I need you to go and find the people whose names are written on these big pieces of paper and I need you to go and communicate with them to make sure that they come to your doctor's office.

So, I'd set them up as if we're playing doctor, right? And I'd say, “Hey, I'm gonna give you lists of all the patients, all the records.” Then, I would take a big stack of 20 pages, and I would drop it on your desk. And I would say, “Okay, go contact all these patients.” And then as soon as you start pulling off the pieces of paper reading which patient, I would drop another 20 pages. And then, when you start actually calling the patients, I would start calling you and have other patients call you, and then I would drop another stack of 20 patients.

And the reason I’d say that is because the fundamental process here is that you are receiving your patients via paper, via the fax, via email. And because of that messy process of me dropping papers on your desk, you're left with this really painful process of having to go and read through the documents, understand what's going on, and contact these patients so that you can get them care.

A lot of people may say, “Hey, Trey, stop dropping paper on my desk. That's a pain for me to deal with.” And to that, I’d say I don't care. It's really easy for me to drop pieces of paper on your desk. I'd like to keep doing it. And so there, you have a problem.

You can't really convince Trey to stop dropping pieces of paper on the doctor's desk. But the doctor really doesn't like receiving that paper. But they're willing to do it because it's the only way they get patients. 

And so we basically come in and we say, “Hey, we'll take care of reading that paper for you.

We trained computers that can basically understand the messy paperwork that's coming in and automatically do the work needed to get those patients in, make sure we have all the information, and schedule that patient. And because we can do it instantly, you're never getting backlogged with the paperwork, with the documents.”

This is an interesting opportunity in that nobody has really come along and said, “Just let them drop them, just let them keep doing it.” One of our advisors is a highly-regarded surgeon and he's said this industry is littered with the bones of startups that have just begged and pleaded for people to go use their alternative system. Meanwhile, we're like, “Just keep using the damn fax.”

What excites you most about your job?

I love our customers. I think they have they have an awesome winning story. They get to grow their practice, but they're not actually who I'm always thinking about when I think about what it feels like to solve this problem. I think about the other two sides. So, not the recipient of the documents but actually, the patient and the sender. 

A lot of times, these senders, even though they're sending you documents via fax, they actually really do care that whoever they're sending the records to gets it and they want to know that it all worked out They want to know the patient was quickly seen. That's why they're sending you their business. They're sending you their patient—that's who they care about. 

The patient also wants to be seen, but we don't want to have to rely on super high agency patients to make it happen. When that record is sent, they should be contacted immediately. So one, the doctor sends off their information and just has complete and utter confidence that their patient is going to be nearly instantly contacted and handled as well as they would handle them.

And then two, from the patient's perspective: When, for example, my co founder is discharged from the hospital for sudden really bad gastrointestinal issues, he's contacted 20 minutes after walking out while it’s still fresh on his mind, so that he can book his appointment and not have a three-month runaround cycle of trying to go and get booked by the time the symptoms aren't really there anymore.

Which trend will change the future of medicine? 

I think so many providers, and the younger ones especially, are waking up to the benefits of having open APIs. The reason we exist though, is because we have this insane integration engine that allows us to get into on-prem, into old file storage systems, into systems built decades ago. But my God, our solution can be so much more efficient to implement and so much faster when we work with really cutting-edge EHRs that are very API friendly, which allows us to focus on what we like to do best, the document side of things, and allows the EHRs to do what they do best, which is the storage side of things, the record management, the clinician-facing side of things. 

Looking back, which trends have you missed or underestimated? 

From a technology perspective, people getting excited about specialized models built for singular purposes within healthcare.

I think one of the reasons AWS sent a film crew to interview us when they saw what we were doing was because so few people are actually saying, “Hey, what if I built a model for this one particular problem, for this one particular thing?” And we've seen this in radiology, like what happens when you have really good models built for particular tasks.

But I think, ironically, the popularity of these very large and generalized models like ChatGPT has woken everybody up to the idea that, “Hey, wait, actually the people creating really specialized models for specialized tasks are actually making leaps and bounds as well.” But a lot of people weren't looking at that, because so much of the noise around this technology really came from people just doing basic statistics analysis and not not actually using real ML techniques.

Which MedTech initiative or startup deserves more attention? 

If I looked at med tech startups, I’d talk about Deep 6. They're in the space of dealing with unstructured data within EHRs and trying to do useful stuff with it. It's one of those use cases that we're not really well positioned to solve, frankly, but that I think is just such an obvious need.

If we can read through notes and determine who would make for a great clinical trial candidate, for research studies, that has nothing, frankly, to do with the information flow of unstructured data, the way that we live, but it has everything to do with where healthcare is really at today—taking the unstructured data and then doing something about it, which in this case is helping life sciences companies get more data to develop cool, interesting therapies.

Where would you put a million dollars? 

I live in New York, in the heart of Manhattan. And I don't get nearly enough sun or nature.

So, I'd figure out a way to build another another park. But it would be a different type of park. I would really want to build a mud room.

I grew up in California. I was a huge nature buff. I like surfing and hiking on the weekends. You can't really do that in New York. It’s part of the reason it's a little bit easier to work really hard in New York, I found. And so, I actually wanted to start the company out here and not be so surrounded by tech. That said, the idea for the mud room is basically: What if we could take a rooftop somewhere and get a bunch of soil with the perfect bacteria in the soil with a bunch of good plants and you could just walk around that barefoot? I just feel like that would be good for a lot of New Yorkers. I don't know if a million dollars would do it, but I want good soft mulch I can walk in barefoot. 

What's the best advice you've ever received? 

I would say don't sell anything you wouldn't buy yourself. I completely ripped that off from Charlie Munger. It's something that I really try to hit hard with our sales team, which is that I have no interest in selling stuff where people are going to feel like we got the best of them.

I want to sell stuff and feel like, every single time, this person is getting an incredible deal. We're going to make a massive impact on their business. We're going to make a massive impact on the network of businesses they work with. It often just feels like basic empathy building and it has a bunch of good ripple effects that apply really well to business overall.

Also, stop saying AI. That's another thing I would say. I hope you can tell we don't say it. I don't say it. I think it's one of the most cheap marketing and sales tricks that people use. They show some tech and then they hand-wave over it all and say, “It's all AI-powered.” It's such BS. There's so much smoke and mirrors in it. I gave a talk on this at our last conference, and I think it actually was really well received because I was like, frankly, most of the time you hear “AI,” you should just replace it with “BS.” Because that's what most people think. It's usually just a way to try to help get a deal cycle through so somebody can sell it to their boss or what have you. Of course, there are very good applications for very good models. But you should always understand what the model is going to do, and what the model grid is. And I really think one of the things we try to do—and hopefully it's helped us stand out—is just not saying AI. And I always say I would scream from the mountaintops to get people to stop doing it. So I will do that now.