Jeff Sorenson started his career as an imaging field engineer more than 25 years ago and today serves as CEO of TeraRecon, the leader in medical advanced visualization and artificial intelligence solutions. After one discussion, this is what we know about Jeff Sorenson: he loves what he does, he lives in his work, he remains curious at all times and he believes in being dissatisfied with last year’s work.
The enemy of innovation is pride – being too attached to the work that you’ve already done, which was good work. Pride creates a blindness to the work that you should be doing.
Jeff has a clear vision for TeraRecon, consistently reflecting on learnings – product, culture, partnership – as he focuses on what he calls the white space opportunity, the realm of the possible that is morphed into a completely actionable plan. He shared a little context of each with us to provide exclusive insights to the Healthcare Performance Insider reader.
(Spoiler alert: Get ready to take notes. We’ve already suggested he write a book. You heard it here first.)
It all starts with the product.
Jeff is clear here – customers are trading their money and their time. They have a limited number of things they can focus on so your product has to have the highest impact or most important thing that can deliver the biggest results. And focusing on the product involves understanding your customer.
Customers don’t always know what they need. It’s very comforting to say the customer is always right. I think the customer is always right when they’re articulating their problem but they’re not always right when they’re articulating the solution, roadmap, or tactics and strategies that could best solve that problem. That’s why you see a big problem between translating wonderful ideas into real products. Often times great ideas really can’t be used in the creator’s own backyard, in their own hospital, in their own health system, on their own patients. And that’s because there’s a big difference between an idea, a technology that works, and a product.
If you look at Tesla, who’s delivered the first mass market successful electric car, they didn’t spend any time thinking about who their competition was or how they could do a partnership to get their hands on a car like everyone else’s and electrify it. They started with the foundational idea that we are building an electric car. And then when you understand the consumer’s needs and the logistics around what would make a successful car, you design it completely differently. It’s got to be lighter; it’s got to be more aerodynamic; it needs to be more automated. There’s lots of things that would design a new customer experience for an electric car that really is nothing like the customer experience for a non-electric car.
TeraRecon is known to have a strong company culture. Employees post great reviews and if you talk to them directly (which we have), they feel empowered, supported and driven to be better. But how does this happen? Jeff doesn’t believe for a second that it’s all him. He believes culture is mostly people driven. TeraRecon is 100% a software company; they have no shipping department or product physically going out the door. Their software is built by people – hands on keyboards, brilliant minds putting their work into a collective work product. The ability for people to truly be their best or even to truly be their worst – 10% bad, 90% good but move very, very quickly is the real culture of TeraRecon.
If you look at the most innovative companies that are on the planet today, they are companies that started organically. When you have a small company, and it grows into a medium sized company, that grows into a medium-large company, you’re going through a lot of growing pains. People talk about wearing many hats. That’s hard. But in a way, many hats in a very small environment is easy and necessary. It’s easy to understand how to do that. But when you’re growing very fast, you have to understand how to have people share hats. You have to understand how to give up hats. You have to understand how to drop hats if they’re not the most important hats. And all of that can be very stressful. And you’re asking a lot of people that are not only outside their comfort zone but actually sometimes not capable of doing what’s being asked of them. You still need them to try, you still have to accomplish things as a team. People who see a hat that’s dropped need to pick up that hat and there needs to be no discussion about why it dropped. It’s about a team focus, a team objective, a team culture.
Jeff presses on, noting how becoming a market leader brings a whole different set of challenges. All of a sudden you have to be highly responsive to tens of thousands of customers while still being innovative. He stresses repeatedly though, no matter what the challenge, if you get culture right it’s very easy to get everything else right. But if you get culture wrong, it’s impossible to get very much of anything right.
Make friends. Partnerships help us all.
In healthcare, you have technology systems that don’t communicate and customers trying to build products out of systems that need to connect. Jeff shares the word coopetition with us – then explain further.
I started to realize pretty early on that “coopetition” doesn’t work. Having a bunch of companies that compete to serve the needs of a customer and somehow need to work very closely together to keep the customer happy, all gets in the way of the customer necessities. It just doesn’t work.
Our customers are looking to buy solutions today. They need things that they can talk to their peers about, go look at best practices from other healthcare facilities, know what they’re getting, and trust that the vendor or the group of vendors can actually deliver a complete solution. Not a kit, but an out of the box, proven, high ROI, high patient impact, easy to install, easy to administrate, complete solution. Just like the Tesla – you didn’t just need a car, you also needed car automation technology and a series of chargers around the nation or the thing just will not work. The same is true today in healthcare. There’s a new standard of excellence.
We need to have providers and vendors of different types work together and deliver complete solutions that make sense, that are proven out of the box. That’s the new definition of a partnership.
He shares two examples of current TeraRecon partnerships that are having an impact on the business – and patients.
- A perfect example is the relationship we’ve built with 3DR Labs. 3DR runs the tools to make the great 3D volumetric imagery and measurements for very complex procedures. They provide people that are highly trained with very highly optimized workflows to make sure that the customers get exactly the work product they need. TeraRecon is focused on delivering the tools to 3DR that can be used to process the images to deliver results. We share the same customer.
- Another very interesting partnership happened when we looked at our capabilities and said: we’re the world’s best at what we do with volumetric image processing, but what do we do for imaging data? What do we do for non-imaging data? What do we do to apply AI to data? So, we started looking for partners to become part of a bigger ecosystem. In March, we sold to Symphony AI, a portfolio of AI companies that have the types of technology we need to process vast amounts of big data. Now we’re partnering with a single unified vision.
Jeff describes what the first images of the disease looked like: a wet paper lunch bag that had been crumpled up and put into the lungs. Very dense and clearly abnormal looking. His immediate response was to shut down all engineering and start working on the problem. The images were breaking the software and they needed to optimize an algorithm that could allow providers to calculate and quantify the volume of disease in a cloud environment, in the electronic health record. He highlights two critical learnings over the past many months.
- Too much of our knowledge and expertise sits at the hubs hospitals of a health system. When a healthcare emergency happens, patients flood into the spoke. The difficult work, the challenging cases are then sent to the university medical centers and large hospitals downtown.
A more distributed model would accelerate time to treatment.
- It takes healthcare too long to bring innovation to market. While this is certainly not a revelation, the pandemic inspired TeraRecon to accelerate how they bring tools to the point of care using their already completed AI work to prototype and deliver completely automated AI powered 3D imaging results.
The challenge was fantastic for us. In just over 30 days, TeraRecon delivered everything needed. And again, that comes down to culture. It was 18-hour days, seven days a week for everybody in our company to deliver the solution required. In the end, it was quite a transformational experience.
Thank you, NEXT.
TeraRecon is now focused on simplifying everything. Defining a whole new frontier of artificial intelligence that can develop image processing tools 20 times faster and be delivered to the point of care, right to the bedside of the patient, across many different specialties. Jeff is clear on one thing – TeraRecon will be the conduit to delivering automated results that are actually impacting the patient.
We can put more intelligence in the image processing, and that means that humans need to make less clicks. And when humans are working within our tools, they can do the equivalent of 20 minutes of work in perhaps one minute or five minutes, depending on what the task is. We don’t believe we should be using artificial intelligence to replace the humans in the clinical care paths that the tools are impacting, but instead we should look at it more like a miracle two minutes. We’re creating tools that will save each radiologist time, free up two minutes times 10 in the course of a day. We’re accelerating clinical decision making with images and these tools will be completely productized by the end of 2021.
TeraRecon is the innovative spirit behind much that goes on in advanced visualization and the reason that advanced volumetric image processing is highly automated, highly anticipatory, and integrated into PACS environments that radiologists interpret. They stand alone as the market share leader and we commend them for their dedication to excellence, and the patient.
TeraRecon is once again redefining advanced visualization, now through AI.