InsuranceAdvanced Analytics & AI
Accident prediction project
When a claim goes to court, the probability of losing trial is high. We evaluate strategies which appliance will predict and better manage this accident cases.
- Industry
- Insurance
- Solution
- Advanced Analytics & AI
01 / 03
The challenge
- When a claim goes to court, the probability of losing the trial is high, and this implies a high cost.
- The insurance company needs to predict these cases to apply strategies to better manage them and avoid going to trial.
02 / 03
What Tekne did
Objective
Work the data to be able to generate:
Models
Profiling
RFM
360º Vision
Analytics
Data Delivery
360º clients' vision
Claimant's variables
Policy's variables
Provider's variables
Claim's variables
New variables created
Models
Construction and implementation of a Machine Learning solution for:
- Predicting which cases are more likely to end up in court and be able to manage them better
- Identifying key variables
- Micro-segmentation of customers based on their characteristics
03 / 03
Results
- It was possible to identify the customer segment with a high potential for legal action
- Judiciality was reduced by 13,2%
- It was possible to profile the company's clients
Initial Situation
without model
21%
of claims ended in court
Final Situation
with model
13.2%
of claims end up in court
More cases
View all casesRelated cases
Let's talk
Book a meeting