Reimagining Insurance Operations with AI-Native Automation

Gia Snape: Hello and welcome everyone. Thank you for joining us today for our webinar, Reimagining Insurance Operations with AI Native Automation. In today’s fast-paced and rapidly evolving market, insurance companies that rely on outdated legacy systems and sluggish BPO models face increasing challenges to keep up with rising customer expectations. Experts from Hyperscience are here today to tell us how insurers can unlock new efficiencies and competitive advantages through AI native automation. I’m Gia Snape, news editor at Insurance Business, and I’ll be your host for this session.

Gia Snape: In a world where customer expectations are higher than ever, inefficiencies and back office operations aren’t just an inconvenience, they’re a direct threat to staying competitive. Legacy systems and manual processes often lead to poor data quality delays and costly operational models that prevent insurers from making smarter, faster business decisions. But there’s good news: advances in AI native automation are transforming the insurance industry. Today we’ll explore how Hyperscience’s platform addresses these challenges, enabling clean, accurate data extraction that drives ROI, streamlines workflows, and improves decision making. To help us reimagine what’s possible in insurance operations, we have two incredible industry experts with us today. Brian Weiss, CTO at Hyperscience, is a longtime technologist with extensive experience developing deep expertise in structured and unstructured data. We also have Rich Mautino, Director of Sales Engineering at Hyperscience, who leads the pre-sales organization for North America.

Brian Weiss: Thanks, Gia. We’re glad to be here. Rich does look more and more like James Bond every time we do one of these. What we wanna do today is talk about some key trends that are impacting insurance operations and where we see those going in 2025. That comes from respective of how we invest in technology and what our current customers are doing. We’re gonna talk a little about our approach to an AI-led modernization of insurance processes. Rich is gonna do a demo, and then we’re gonna get into some recommendations.

Brian Weiss: Hyperscience is roughly a 10-year-old company, a pioneer in AI and ML-led automation. The company was founded by data scientists and ML engineers roughly 10 years ago who turned their attention to the hard problem of human information inside the enterprise. How can you get a machine to understand like a human what it’s looking at and efficiently and accurately move that through a business process, effectively creating a digital worker inside the enterprise? They realized that the only way to solve that problem is through machine learning; you can’t apply a rule or a set of rules to try and understand something as complicated as handwriting. They did this long before BERT existed and way before any transformer papers. The idea that you could use deep learning to bring a human-like understanding to the problems in the back office was the founding principle of Hyperscience.

Brian Weiss: We’re pioneering the concept of hyperautomation with AI by building sovereign models with customer data. What you’ll hear about is the platform we bring to the market, which allows our customers to use their own data along with Hyperscience to build high performance models for automation—99% accuracy and 98% automation. That’s a huge shift from traditional technologies that have just used blunt force to get it done.

Rich Mautino: We have a strong track record in the insurance industry. Some of our customers include underwriting, claims processing, and enrollment. Those bread and butter functions of the insurance industry have a lot of human information embedded in them and a high amount of variability that’s very hard to systematize. Those use cases end up being spots where companies move their automation rates from 70% and accuracy of 60% with lots of BPO spend to 99%. Some of these use cases are cracking open even more for us now: claims management and fraud investigation, which are really about long form. You’re not just looking at a key value pair, but looking at understanding what’s in a 50-60 page document narrative, maybe looking at understanding the patterns that might indicate fraud.

Rich Mautino: Let’s talk a little bit about the 2025 insurance trends and priorities specific to AI automation. These processes rely on billions of documents every year. Oftentimes there’s a lot of different goals and functions ranging from onboarding to auditing, underwriting, all the way to fraud investigation, claims processing, and then renewal and endorsement. The end result is that half of the business processes out there today are still paper based. Ultimately 90% of the data that our customers show us is unstructured and not machine readable. It presents great challenges to taking it from current state to future state.

Brian Weiss: If you look at this study from Accenture, there’s $170 billion of premiums at risk from customers who switched because of bad experiences. If you take more than three to six months to get claims work done, your risk of losing that customer increases. So there’s this combination of needing to be able to cut costs and be efficient, but the key factor is what’s happening at the top line? Am I driving customer adoption? Creating a customer experience means faster claims processing. So where are my manual processes? How do I automate workflows that are stuck?

Brian Weiss: Automating workflows requires operational efficiencies. Errors are incredibly important to look at because once they get further downstream, they get compounded. The criticality here is high data accuracy. You need to think about how you’re getting data and the accuracy of the information that drives it. We are right now at a transition point in technology where the opportunity to shift from rules-based approaches to a model-based approach that you can actually train yourself where you have visibility to the accuracy is where you are gonna unlock GenAI or potential. Using AI to do what people do in a controlled manner to drive high accuracy is what gives you compelling customer experiences.

Brian Weiss: The origin in this industry of the technology is document centric. What that means is that you’re taking anything on the left and you’re trying to match it to a template. You will never chase the variability of the world by matching yet another piece of variant here. You end up with your pass/fail exceptions, and then of course downstream your SMEs have to catch the exceptions. The amazing thing about the legacy approach to this is you have this black box that thinks it’s right. It’ll punch out 90 out of a hundred documents, but if you read the documents that came out, they’re wrong. There’s no accountability for accuracy in those systems. And so what you end up with is this assumed accuracy, which of course isn’t there, and then you need to dump the whole thing out to a BPO after you’ve built your own accuracy harness to try and figure out if it got it right or wrong.

Brian Weiss: There’s a new way to think about this, which is you take machine learning at the core of it, and you train models plural to do what people do when they look at a document. “Hey, what is this? Here’s a box of documents. Classify ’em for me. By the way, take the ones that you’ve got and I need you to find these 15 pieces of data, but they might live anywhere. Just go find them. And once you found them, what I want you to do is actually read the handwriting and tell me what’s in there and get it right.” What Hyperscience is doing is giving customers a platform to build and train these digital workers on their own data. You start with our quick know, and you build up these models which become incredibly accurate. Not only are they accurate, but they’re also accuracy controlled. You have the ability to make them better.

Brian Weiss: The other thing that we’ve done at Hyperscience is change the approach to humans and machines. We’ve taken a just-in-time human-in-the-loop approach to accuracy. Instead of the black box dumping it out to you and you figuring out if it’s right, and when it’s wrong you pay a BPO to do that work, the model—think of it as a digital worker—will actually raise its hand when it can’t meet your accuracy standard. “I’m a little bit confused between the A and the U here. I’m gonna ask a human for help on this one spot.” By unclicking that, you use a fraction of the labor that you would spend on a BPO to do the whole document to sit next to the machine and unstick it in that very specific place. Changing the mix so you’re thinking about how humans and machines interact to get an outcome like 99% is really what Hyperscience is driving.

Rich Mautino: Ultimately all of this is exciting, but if there’s not results, it doesn’t mean anything. With Legal & General, we started with them five plus years ago to streamline their automation in the insurance workflows. Back in 2019 when we came on board there, their customer experience they say has never been better, and as a result, they’ve won the Moneyfacts Consumer award the last five years in a row as the life insurance provider of the year. With the VA, processing billions of pages, the VA secretary is quoted as saying that processing has never been faster. Those claims are the fastest in history. And then finally with Corbridge Financial, what that resulted in was a 70% reduction in data entry time that took them up to 95% accuracy in handwritten forms. And that’s coming up from 10%. It’s over a thousand percent ROI when they were looking at 10 million.

Brian Weiss: Gartner will tell you models plural are the future. It’s not really about there’s gonna be one magic model that you can plug into that’ll fix it. The future is about a portfolio of models. You don’t need a billion parameter model to get the square root of 64. I can get you a very efficient answer with a small language model that’s built on your data. And then I can go ask that model really tricky questions like, what’s this customer’s intent? This concept of using AI in an ensemble portfolio approach is absolutely the future. Otherwise, what that means is model ops now becomes very important. Models are tools, a platform that allows you to orchestrate these tools in a way with visibility, transparency, and continuous improvement is where the market is moving.

Brian Weiss: I want to ask you guys, don’t click the “combination of the above” button on the poll. Try to balance like, on balance, what do you, where do you spend more of the work? Machines or people? I am interested in how people think about the balance between cost savings and customer experience. Let’s look at where your people are doing work that you don’t think they need to. Are they on simple or on complex documents? Look at where you’re spending money on people. Where is the friction? Where do you spend money on a BPO? What slows your process down and start there. Hyperscience is one of those applications that CFOs absolutely love because you can take the AI budget and do something transformative that drops ROI to the bottom line immediately.

Rich Mautino: I’ll hop into a demo. This is all in the browser, so you don’t have to be a data scientist to use this. This is meant to be something that can enrich your orchestration of processes. Imagine an insurance claim process where all of these documents come in. Historically, they get scattered to the wind. Here, what I’ll do is I’ll simulate that coming in here, and when that submission comes in, it’s gonna hit the system and you’ll see it actually start running automatically. What I like to really emphasize here is watching the flow run. Our blocks and flow architecture is what I think sets us apart. What you’re seeing here is the true orchestration.

Rich Mautino: We’ve got a task ready. This is a human in the loop function that I’ve actually intentionally created. It’s identified two documents here. It’s still got some unmatched pages here, and it’s raised its hand and it said, “Hey, I need some help with these nested tables here.” So rather than having you do it from scratch, it’s going to show you what it thinks and give you an opportunity to either say, yes, I agree, or no, let me help you out here. In this case here, you can see this complicated nested table. It’s actually done a really nice job with it. It’s got everything correctly coded so we can say, yep, good job.

Rich Mautino: Now you can see we’ve got a task ready. This is asking for some help with transcription. In this case here, you see it is flagged at postcode and it said, “Hey, I need some help on this. I’m not positive.” And you can see we’ve given it some intentionally some very tricky information here. Is this a five? Is it an S? Is that a one? Is it an I? What I’ve done for the purpose of the demo is actually trigger this. And in this case they’re gonna correct it. That’s a zero. But the neat thing here is that we wouldn’t actually hit this in the real world because we could either set a data type here where you can see I can actually tell Hyperscience what to expect. I can say, “Hey, it’s always gonna be letter, letter number, number for example.” Or use an API to check if it is a valid postal code.

Rich Mautino: Now what it’s done is loading this admission, starting to build a shell. It’ll grab full page transcription, and then you can see there’s a lot of other orchestration layers that it’s gonna move very quickly through looking for entity recognitions. “Is this a clinic that we have on file? Where’s the policy number?” It will call it things. It’ll check for validity of policy. And then finally we’ve got a fork in the road here: is additional human validation required? Depending on what your needs are, you can set it up to either say, “Hey, I have everything I need to make a decision on my own,” or it can actually raise my hand and say, “Hey, I need help. I need some guidance.” Because making a wrong decision in the insurance world is very catastrophic.

Rich Mautino: The supervisor has got the ability to quickly review all key information and they’ve got that sort of AI co-pilot experience here on the right. This is what we call custom supervision. If I need to make a decision rather than having to go look for everything, it’s all where I need it. In this case here, I got my policy number. Looks like the policy started back at the end of March of 2023. Here’s the patient name by the way. Looks like they’re a little unhappy. We’ve got some negative sentiment here. And what it’s doing is making a recommendation and telling us why. So in this case here, it’s saying, “Hey, this claim should be rejected.” And the reason why is that the first symptom of this actually occurred before the policy start date.

Rich Mautino: Furthermore, it’s taken a very lengthy doctor’s note, which sometimes would need to kind of be looked through. In this case here, it would just summarize that for me. So I have everything I need here to make a decision. And what I can do here is click reject the claim. I can put the notes in here. So in this case here, now this can go in a vector database and in the future, if there’s similar claims made to this, we can expedite the decision and also get to that more accurately here.

Gia Snape: Let’s start with the first question. How does Hyperscience handle complex documents like multi-page forms or handwriting?

Brian Weiss: Actually that’s bread and butter stuff for Hyperscience. People often think like, oh, great, I’m gonna get a document and it’s gonna work fine. But no, actually once don’t you get a 50 page document with a nested table that’s split across three different pages that changes? We ship 35 plus models that are under the covers like Hyperscience does all of that work. Very, very good at that kind of computer vision management of what’s coming through. And handwriting is difficult. It’s hard. We have been doing deep learning ML AI on our handwriting models for 10 plus years and continue to refine them for all of our customers. So we are very, very good at extracting handwriting correct. Whether it’s scribbles on a page or whether it’s inside a bounding box.

Gia Snape: Does your solution easily integrate with existing claim systems?

Brian Weiss: Yes. We wouldn’t have the customers we have right now without being able to put inputs and outputs to the systems, and also the data that’s required to make decisions. Whichever systems that lives into, we’ve been around for 10 years, we have some very large customers with complex systems. So when it comes to integrating, moving data in and out and through Hyperscience or bringing data into Hyperscience from systems to make a decision, yeah, we have a robust set of capabilities there.

Gia Snape: And what level of accuracy is Hyperscience seeing with unstructured data extraction?

Brian Weiss: So when you say unstructured, if you mean like just fully scribbled handwriting and things like that, we can get up to 99%. And this because we’re constantly fine tuning that content. The places where it gets interesting for unstructured is we actually see more of kind of slightly structured long form documents are where things are getting really interesting. For example, a credit swap agreement. This is 50, 60 pages and none of ’em are the same. And much of the data is delivered narratively. We’ve helped a number of customers solve that where they were starting like, “Hey, I’m just gonna throw it at an LLM and tell it to go.” So what we do is we actually chunk that data up and so we’ll section out which pieces are what, and we’ll basically pre-annotate a long document sometimes into a vector database, which gets even more interesting ’cause you’re accumulating all that interesting data.

Gia Snape: Could you provide insight into how your machine learning models continuously improve over time, particularly for handwriting recognition? For example, in cases where a date appears outside the bounding box, how does the system learn from a single correction to avoid repeating the same issue?

Rich Mautino: Bounding boxes aren’t an issue that we typically even see it fail on the first time. And that’s just by the way we train the models to begin with. But fine tuning is how we’re constantly making small tweaks to the system to make sure that, you know, if this person writes their Q a little funny, for example, that it knows what that is and gets it right in the future.

Brian Weiss: One of the things we have built into the handwriting extraction is built-in fine tuning. And so what that does is it does the statistical analysis with some QA after the fact about when the machine thought it was wrong, but it was actually right. And so what we see is a lot of times as you process data through, you end up calibrating when the machine thought it was wrong but was actually right. And so we’ll typically see a fine tuning accuracy bump in specific to handwriting extraction of 10 to 15% with a fine tuning process. That really runs automatically as part of what’s built into the solution.

Gia Snape: Wonderful. And that brings us to the end of today’s session. A big thank you to our speakers for sharing their expertise and to all of you for joining us. Don’t forget to scan the QR code or click the link in the chat to learn more about Hyperscience and everything we discussed today. Thanks again, and we hope to see you at our future events. Have a wonderful day.