TOUCHLESS CLAIMS | 6
Lemonade utilizes a machine learning-driven bot called "Jim" that serves as the "first notice of
loss" for 96% of claims. Jim can uniquely manage the entire claims process -- minus human
intervention -- and has been known to pay out settlements in just 3 seconds.
Jim's 3-second settlement set a world record. During that short timeframe, he
reviewed a claim for a stolen Canada Goose, cross-referenced it against the
user's policy, ran 18 anti-fraud algorithms, approved it, sent wiring instructions to
the bank, & delivered the good news to the customer.
In 2019 alone, Lemonade's Jim handled over 20,000 claims and disbursed $2.5 million all by
himself.
Jim is powered by powerful machine learning algorithms which enables this bot to readily
identify the nature of a claim and its severity. If a user files a claim in an area with a hurricane,
flood, or wildfire, he takes them through a special, quick flow process addressing the
emergency, so the customer can spend less time filing a claim in their hour of need.
Claims Segmentation
It is estimated that up to 40% of the claims cycle time is wasted by manually assigning and re-
assigning the claim - assessing the complexity of the claim, identifying appropriate resources
(including loss adjustors) and adjusting workloads and schedules.
The utilization of predictive analytics and AI algorithms
can help segment claims by complexity and severity -
leveraging the company's business rules and
augmenting them with machine learning tools that
consider all relevant historical claims data, as well as
recent claims experience with regular updates.
The less complex (and less risky) cases can be directed
to a self-service application that can speed the process
along without human interaction. The more complex
cases can be steered to an adjustor or claims handler -
especially if they include bodily injury and medical
assistance or offers a higher litigation risk.
