After a hurricane hit South Florida, a mid-sized insurance company accumulated a backlog of 4,000 policies that needed to be rejected within 90 days or they would be automatically renewed, sometimes at substantial loss to the business. To get through this backlog, the team would need 10 additional FTEs.
Instead, they worked with Invisible and burned through the backlog at 6x the normal pace. Since, Invisible has run up to 14 processes for this business and seen them through another hurricane.
Invisible mastered an 80 page workflow that prepared policies for the internal expert team. Not only was the team able to burn through the backlog of 4,000 policies, they were able to review them an average of thirty days earlier than their previous best. Not only was the team able to make decisions more quickly, they were able to make better, more rigorous decisions, as this additional time allowed for further rounds of review.
Before Invisible, underwriters spent a lot of time gathering basic information and making simple decisions. For example, to underwrite housing policies they had to:
Those decisions are all simple for someone else to do, hard to teach a bot to do, and a sub-optimal use of underwriters’ expertise. While large companies use RPA combined with AI for these types of problems, the solutions that work aren't affordable for mid-market companies for whom cashflow is a concern.
Working off of an 80 page and highly detailed standard operating procedure, Invisible created a reliable series of workflows that helps underwriters optimize their time. It divides work meaningfully between trained people and simple automations.
If you've never worked in underwriting before, it may be hard to visualize these sorts of simple decisions. Here's one example: To validate certain housing policies, one must verify that the number of stories listed on building records matches photo evidence. It turns out that the human eye is still more efficient at this cross-checking than computers are.
Not only are humans faster at making sense of photos than untrained machines are, they are also better at handling exceptions. For example, you can see in this workflow that agents are equipped to apply judgment based on a handful of scripted exceptions.
In the insurance industry, the core ability that's often needed to scale processes—in addition to working with legacy systems—is to scale simple judgments. Invisible's unique way of managing the relationship between people and machines enables our clients to do just that.
Invisible creates cost-efficiencies for this business every day and provides critical surge capacity when the region is beset by natural events. As a result of our partnership, the company can grow efficiently in good and bad weather.
Since, Invisible has helped nearly every team in the company gain efficiencies, from processing E&O documents, to automating claims intake, to QA'ing IT tickets. Most recently, a competitor went out of business due to (another) hurricane. This time Invisible helped the company increase its capacity by 25% in one month, and they were able to meet the new market demand.
The insurance industry uses automations heavily, but affordable automation solutions are not well suited to workflows that run on unstructured data and which require decision-making. These processes can be applied to insurance, increasing efficiency where it's hard to gain.