A few years ago, Masa Kato, a data scientist on Ranjini Vaidyanathan’s team at CCC Intelligent Solutions, approached her with an idea.
He had discovered a way to digitally capture different views of a damaged car and automatically identify whether it was the left front corner, the right front, the rear, etc. — a valuable piece of technology for the company, which aims to digitize the auto and casualty insurance industry.
But Kato wanted to take this tech one step further.
“What if we compressed the neural network and made it work really fast?” he told Vaidyanathan. “What if we made it run on mobile so that as the user walks around the car, they're able to automatically capture those images instead of having to take a bunch of photos manually and upload them later?”
Vaidyanathan loved the idea. “Let’s work on this,” she said.
As it turns out, it only took a couple of weeks to push the idea forward, thanks to CCC’s NXT Lab. The company’s innovation hub brings together teams across product, technology, data engineering, data science and even its customers to imagine and experiment with new advanced-technology solutions on the horizon — or, more aptly put, that are next.
Kato and his team met and discussed the idea with the mobile development experts on CCC’s engineering team. Everyone agreed it was doable, so they experimented, collaborated and went through a trial-and-error process to develop a working prototype. Within a few weeks, the team was able to demonstrate their proof of concept to a customer.
“How wild is that? The gap between somebody proposing an idea and demoing it to a customer was just a few months,” Vaidyanathan said. “NXT Lab made that accelerated process possible.”
Some time after, CCC was even able to patent the technology.
The importance of CCC’s NXT Lab is no hyperbole. Just down the stairs from the main entrance of CCC’s Chicago headquarters is an entire floor dedicated to innovation.
“If you talk to our CEO, he’ll say our innovation hub is what’s going to take us forward,” said VP of Product Management Yury Pensky. “And we're capitalizing on it.”
What CCC Does
NXT Lab gives you an opportunity to work directly with customers. What’s that like?
Associate Director of Data Science Neda Hantehzadeh: When building AI solutions, for example, we build models of general applicability using broad datasets reflective of the industry as a whole, with configuration options for customers to ‘tune’ on their own to their business needs and processes. It’s exciting to see it all come together.
Associate Director of Data Science Ranjini Vaidyanathan: Learning about our customer’s needs is fascinating. I was recently in a customer meeting, and we were talking about our capabilities. At one point an executive from one of our customers shared their pain points and what they wanted to solve. They wanted to make the consumer-facing experience simpler. Being in that meeting and learning the customer’s pain points was so insightful and motivating.
VP of Product Management Yury Pensky: Simply put, it’s a privilege to work directly with our customers. When we collaborate, we deliver solutions that delight them. Working at CCC and with NXT Lab means face time with customers and critical inputs to guide next-generation innovation.
One of the most successful NXT Lab products is CCC Estimate – STP. Tell us more about it.
Pensky: CCC Estimate - STP is a product that solves a huge need for our customers. U.S. insurance companies spend about $3-$5 billion annually just writing estimates. At CCC, we have a first-to-market solution where we can automate the writing of those estimates. How do we do that? People like Neda and Ranjini work very closely with product teams to determine how the model should work and what AI training data they need, all while receiving customer feedback on these solutions.
It's one of the best examples of what we do in the NXT Lab because of how transformational it has been and will continue to be for our industry. Three years ago, when this concept first emerged, it was a pie-in-the-sky idea, but it was a huge market need with our customers, and by working with NXT Lab in this collaborative manner, we were able to incrementally deliver the right solution.
Hantehzadeh: One of the things that makes our STP product special is the collaboration between AI and decades of historical knowledge from customer rules and settings. This has enabled the product to deliver unique capabilities in creating line-level estimates. And that’s what makes it distinctive for our industry and difficult to compete with.
You get to experiment with innovative technologies like Metaverse, Blockchain and 3D printing within NXT Lab. What’s that like?
Hantehzadeh: It’s really cool to be on the cutting-edge of a lot of advanced and emerging technologies. For example, a few years ago we were working on new technologies within AI that weren’t so popular. Now, they’re broadly used. It takes time for an industry to familiarize itself with new technologies and the opportunities they present. That’s why it’s crucial for us to stay ahead. We’re always up to date with what’s on the horizon in the natural language processing and computer vision space. How will 3D reconstruction change the insurance industry? How will these new, huge language models change the industry for insurance carriers or accidents? We’re tracking all of this and thinking more futuristically — and, most important, demonstrating that to our customers.
“How will 3D reconstruction or these new, huge language models change the insurance industry? We’re tracking all of this and thinking more futuristically — and demonstrating that to our customers.”
Vaidyanathan: We use the collaborative nature of NXT Lab as a way to try new ideas and figure out if the new technology in the industry is applicable to our customers’ data and use cases. So, not only are we experimenting with new tech and finding ways to make it work, but we’re also eliminating the ones that don’t. For example, we were experimenting with detecting damage on images. We asked, “What if we built a multi-task neural network to do that? Would that be helpful to our customers?” And when we saw that it helped, we scaled and expanded it to a lot of different tasks to create a large-scale neural network.
What’s your favorite aspect of NXT Lab?
Pensky: Large companies like Microsoft and Amazon have innovation labs in-house, but most smaller companies have separate consulting business teams associated with their innovation hubs, which tends to slow down the creative and innovative process. You’ll be assigned an account manager who will help you think through the situation. Here at CCC, you actually work directly with engineers like Neda and Ranjini — the people who are actually building the solutions. It’s much more hands-on and faster.
Vaidyanathan: I just love how it’s a creative space. We have a lot of ideas on the research side, and NXT Lab is a collaboration space where teams can talk things through. It helps us get from an idea to something that's a working prototype, and it bridges that gap really fast.