InsurTech Ohio Spotlight with Matt Leiv
Matt Leiv is the Head of Sales at TAZI AI, a global AI company that created a Machine Learning platform that enables business experts (and data scientists) to easily create, update, deploy and take actions with ML. Matt was interviewed by Michael Fiedel, a Managing Director at InsurTech Ohio and Co-Founder at PolicyFly, Inc.
Matt, what sort of gaps in data science are insurers often facing today?
“There's a big range. You've got a combination of large carriers with large data science teams that are having obstacles. Then there’s the rest of the market, the tier-two, smaller carriers and mid-size carriers that have minimal or no data science resources. With a large carrier, you've got teams of dozens to hundreds to thousands of data scientists, brilliant people putting together machine learning models and technologies, but they have a really tough time getting end user buy-in.
They'll create this amazing solution that nobody wants to use. It's putting in all this effort with no real results. That's a really difficult position to be in for data scientists. As a carrier, that's an immense expense that you're not getting benefits from. With smaller carriers, they might be more agile, a little bit more nimble. They can deploy innovation a little bit faster, but they can't get enough data scientists to work on their projects. Data science is blowing up right now. It's a high visibility, glamorous position for a lot of companies. If you're not some big, data science attracting firm, it's hard to get top talent that can do great projects for you. Overall, those are the two situations I'm seeing creating a lot of gaps. It really comes down to resources on personnel and technology.”
What does the democratization of artificial intelligence actually mean?
“I've been speaking about democratization of AI (artificial intelligence) since I first started in AI because we have this great, world-changing technology. That's been a big part of all of the work that I've done within artificial intelligence. For me, it's getting a great, powerful technology into as many hands as possible. You had the first stage where mega companies have small data scientist teams, and they keep growing and building these teams. Then it starts trickling down to where mid-size companies can have data science teams and start deploying AI. Now, there's technology out there. The company I work for, called TAZI, has data science machine learning technology that you don't even need to be a data scientist to deploy in a corporate environment. Democratization of AI is the continual push down of AI from large resource to small resource to, eventually, individuals.”
From a value proposition, what will democratization mean for business owners?
“You have the same thing with software and app infrastructure. You've got a combination of large value propositions, time, efficiency and cost. Then you're going to have to combat that with the question of am I just spending a bunch of time with some cool fun stuff that's not actually getting any results? That's the same thing we see with software and with app development is there can be a lot of movement, but not necessarily any actual direction. With the democratization of AI, I think we're going to see that challenge as well.”
How will data scientists use the additional free time when they're not burdened by more mundane risks from internal stakeholders thanks to these new and available tools?
“That’s a question that originally was scary for the industry because when you bring in all of these different types of technologies, what if an end user has access to AI? What am I going to do? Well, there's still the specialists. They’re still the best people at their jobs. Instead of spending their time across 20 or 30 or 50 different types of tasks, now, they can focus on optimizing model performance and experimentation. It gives them the ability to be more productive, more creative and come up with more beneficial solutions, instead of being so sucked into the huge stack of projects that have to get done. I’m hoping that it gives them an even better life, realistically.”
Is there a case study that you can share about the impact that TAZI has had in implementing this type of artificial intelligence tooling?
“I've got two off the top of my head that are really interesting. You have roles like actuaries where they're trying to find different segments of data within their business. They're trying to find out, for example, what's higher risk or what's more profitable, or they're trying to figure out these segments. How do they price and put things together more effectively and more efficiently? One of our clients told us that their team of 10 using our technology in two days, found more segments than that team of 10 working for 10 years would've been able to find. It's the ability to really get so in depth with your data and understand better than anything that's out there.
Another one is speed. We had a client where the entire sales process from when we met them to when they had insight from machine learning was 47 days. We met them at a conference day, 48 or 47, it was deployed, operating, and they were getting results from it. It absolutely blew them away how fast this technology can be implemented and used. And so, as you're aware, the insurance world sometimes moves a little bit slower than you'd like to. A lot of times projects get hung up on many different barriers. If you can eliminate the speed/time barrier to where the same quarter that something is approved, it’s deployed with results, that is a huge win for the executive level. In one quarter, you can make a decision. Is this something to keep going with or something to cancel?”
What do you expect from the future of AI, and how crucial is it to the strategy and success of insurance companies moving forward?
“If you step outside of insurance, and you look at the S&P 500 (Standard and Poor’s 500), the average time that a company was on the S&P 500 was in the ballpark of 50 plus years. Now, it's around 14, and the same level of innovation is happening across every market. Within insurance, there are a number of carriers out there not just using AI, but using it to great success and having a bunch of amazing results. If you're doing a little bit with it, you're struggling just to keep up. If you're not using it at all, you're gone in 10 years.
Also, on the thread of democratization, you're giving smaller insurance carriers the ability to have an impact that larger carriers can have with the use of AI. It's this chaotic, dynamic landscape of technology implementation that’s delivering tangible results. If you're not using it, and I'd say this to every industry, you don't even have a chance to catch up anymore. You're going to be gone.”