Imagine dining at an elegant restaurant. The waiter escorts you to your seat, you select a five-course meal and then settle in for what may be a long wait as the chef prepares your food. But, to your delight, the server instead brings out each course without delay and at the perfect moment. The food is delicious, and when you receive the bill, the cost is no more than it would be at a neighborhood diner. On your way home, you think about your surprising and pleasant visit.
Groundspeed is that impossible dining experience for middle to large-market commercial insurance submission processing: high quality, fast, and low cost. One of the main drivers is the automation we’ve put in place throughout our human-in-the-loop AI system.
This blog will address the common challenges insurance carriers face and how Groundspeed’s AI platform addresses those pitfalls.
Challenges commercial insurance carriers face
Challenge 1: Complex set of documents
When working at a carrier that serves middle to large insureds, there are many different documents that you need to get data from to make balanced decisions on whether to offer insurance to a prospective customer and how to price their policy. Those documents convey different types of information – exposures, applications, loss runs, and more. Even within each document type, there’s variation in where different data points are available.
To succeed, the carrier needs to balance the speed, cost, data point coverage, and accuracy of retrieving information from those documents. We covered this in the blog post, The End of Underwriting Tradeoffs.
That matters because commercial insurance carriers need to:
- Book more business by having a quick turnaround time for prospective insureds.
- Reduce loss ratios by making decisions using as much relevant and accurate data as possible.
- Minimize expense ratios by limiting the cost of acquiring data.
Challenge 2: Manual Processing
Keystroke-driven approaches can make it difficult to minimize costs and maximize speed, coverage, and accuracy. Whether you do this internally or with a Business Process Outsourcing firm (BPO), having someone manually review and pull out specific pieces of information takes a long time. In addition, it requires paying someone to do what may be inaccurate work. Plus, you’re faced with deciding between buying extra hours to get more information or spending less by cutting corners on the information.
Challenge 3: Consistency
One underwriter or BPO staffer may provide information about a data point one way across dozens or hundreds of prospective insureds; another may do so differently. Imagine if the net total incurred was calculated differently across your organization. Perhaps you don’t have to imagine if you’ve had to manage the fallout from something similar or work to prevent it from happening.
The promise of automation
Where can you turn to when facing all those data extraction challenges? A common answer is automation – take the manual work out of the equation. Automation is always assumed to be faster, and it should drive down costs because you don’t need to pay software programs a wage or salary. When done right, automations are applied consistently, meaning you can get the same broad and accurate data from each document.
In particular, automation can help with:
- Extracting data about each prospective insured from documents, whether it’s about loss history, risk exposure, firmographic data (insured name, address, industry, etc.), or anything else.
- Figuring out the requested coverage type(s) so submissions can be quickly triaged and routed appropriately.
- Classifying documents so you know what documents you have and where you got them is impossible
- Providing additional data about prospective insureds that aren’t in documents but can be gathered from them.
- Packaging all of that data into an easily digestible format.
Where automation can fall short
Introducing automation to your document processing workflow can make a massive impact. So why hasn’t everyone completely automated their data gathering?
- With the wide variety of document types and formats, it is nearly impossible to have an automation that works for every single data point in every document you have.
- There will be missed data points or inaccuracies due to variations between similar documents.
- Creating automations requires a high level of technical skill and insurance industry knowledge – a rare combination.
How Groundspeed addresses those pitfalls
At Groundspeed, we’ve fine-tuned our Artificial Intelligence with a human-in-the-loop element to solve those challenges. While automations classify the large majority of documents and capture the large majority of data points, Groundspeed’s humans catch the ones automations miss. For example, a human can review and fix any data point our automation platform flags as potentially incomplete or inaccurate. We also audit a sample of data points daily to ensure high data quality.
Our expert Document Processing Automation team uses the results of those audits to improve automations and constantly develops new ones. That means our customers don’t need to create or expand their automation teams. You can learn more about Groundspeed’s Artificial Intelligence with a human-in-the-loop system in this previous blog post here.
Groundspeed processes many types and formats of documents, so you don’t have to worry about limited automation coverage. To learn more about Groundspeed’s Loss Run automation, read our blog post here.
Beyond data capture, our automated data pipeline helps commercial carriers get their data quickly, accurately, completely, and consistently by taking this step-by-step approach.
- Optical Character Recognition (OCR) reads documents with high accuracy.
- Predicted and enriched data fields expand data coverage, deriving fields such as geocoded location, NAICS and SIC code lookups based on firm name and address, and VIN-based information on vehicle make, model, year, and gross weight rating.
- Afterward, we map the data fields from each document to a common schema, so that you can easily compare all of your documents.
- Then we automatically package all the data into formats designed for commercial insurance carriers’ systems and send it back to you, saving time for your team.
How Groundspeed can help you
With Groundspeed’s automations and Artificial Intelligence with a human-in-the-loop system, carriers receive data that is accurate, consistent, complete, and quickly and inexpensively delivered from a complex set of documents. Groundspeed enables underwriters to focus on underwriting, and your business can keep loss and expense ratios low while winning more business. Groundspeed’s automations are designed for a relatively rapid launch with any carrier. Groundspeed has captured 6.5 billion middle to large-market data points and processes more than 35,000 pages daily.
If you would like to learn more about Groundspeed’s automation platform, we would gladly show you how we can transform your underwriting process and provide high-quality data quick and for a low cost. So schedule a call with our team today to experience the Groundspeed human-in-the-loop AI platform. We look forward to partnering with your company and helping you unlock the value in your unstructured data.
This blog was written by:
Bryan Quandt – Bryan is the Product Manager for Groundspeed’s internal software applications. He plans the development of automations and human-in-the-loop tools that get Groundspeed’s customers their data as accurately, quickly, and inexpensively as possible.