Real-time Insurance Score Modeling

Making driving safer for all traffic participants.

Yuliya Sychikova
COO @ DataRoot Labs
24 Jul 2019
5 min read
Real-time Insurance Score Modeling
Client Services
Industries
Automotive

Summary

  • Billions of dollars are spent annually on car insurance. While most insurance companies already segment clients based on their accidents history, drivers license score, age and other factors, none do it in real-time accounting for drivers psychotype and live behavior on the road.
  • Our client is building a solution allowing insurance companies not only to update drivers credit score in real-time but also to produce driving-style recommendations to help prevent accidents.
  • With the help of our team, the client built the platform using a simple smart phone and advanced Computer Vision technologies to track and adapt the insurance credit score of drivers in real-time and produce live recommendations.
4
Months
6
Engineers
8+
Technologies

Tech Stack

C++
Caffe
CoreML
Mace
Metal
Python
TensorFlow
TensorflowLite

Timeline

1 Week
Data Labelling and Processing
Data Engineer
2 Weeks
Data Augmentation
Data Engineer
1 Week
Solution Architecture Design
Solution Architect
2 Weeks
Hypothesis Generation & Validation
Deep Learning Researcher
1 Week
Architecture Modelling
Deep Learning Researcher
3 Weeks
Training & Tuning Cycle pt.1
Deep Learning Researcher
4 Weeks
Optimization for Mobile Device
Deep Learning Engineer
4 Weeks
Training & Tuning Cycle pt.2
Deep Learning Researcher
6 Weeks
Mobile App Development
App Developer
Backend Developer
2 Weeks
CoreML / TF Lite Model Porting
Deep Learning Engineer

Tech Challenge

  • We consulted the client on real-time segmentation task which runs directly on mobile phone and achieves approximately 100 FPS.
  • Our team created a customized batch streaming solution to reduce the latency and produce in-batch pre-aggregations on device.
  • Based on the data streamed from the phone, the solution adjusts the credit score in real-time.

Solution

  • By turning a mobile phone camera into a device that tacks the road activity and collects data from accelerometer, gyroscope and GPS, the solution is able to detect cars and track their behavior in real-time.
  • Based on the received data, we built the scoring model that is used by insurance companies to adjust their clients privileges.
  • Additionally, the model evaluates drivers’ behavior i.e. how rational they are, and identifies their psychotype. Having classified the behavior of each driver, the solution is capable of delivering timely reports on drivers' behavior, making gamified suggestions to improve their driving style.

Impact

  • DRL has delivered a working scoring model which is used as a core component of a startup’s MVP.
  • The model not only helps insurance companies to prevent value destruction, but also makes driving safer for all traffic participants.
  • Currently, we are assisting the startup with raising funds and launching controlled beta pilot.

Have an idea? Let's discuss!

Book a meeting
Yuliya Sychikova
Yuliya Sychikova
COO @ DataRoot Labs
Do you have questions related to your AI-Powered project?

Talk to Yuliya. She will make sure that all is covered. Don't waste time on googling - get all answers from relevant expert in under one hour.
OR
Send us a note
Optional
File requirements pdf, docx, pptx

Author

Yuliya Sychikova
COO @ DataRoot Labs
Yuliya is a co-founder and COO of DataRoot Labs, where she oversees operations, sales, communication, and Startup Venture Services. She brings onboard business and venture capital experience that she gained at a leading tech investment company in CEE, where she oversaw numerous deals and managed a portfolio across various tech niches including AI and IT service companies.

Co-Authors

Ivan Didur
CTO @ DataRoot Labs
offices map
Kyiv (HQ)
Max Frolov
CEO @ DataRoot Labs
Tel Aviv
Ivan Didur
CTO @ DataRoot Labs
builds and implement AI-powered systems across different verticals to help our clients operate effectively.
GoodFirms Badge