An AI-driven Approach to Enabling Low Touch Claims Processing

Case Study

An AI-driven Approach to Enabling Low Touch Claims Processing 

Business Objective

Our client, a leading US-based auto insurance provider, wanted to simplify and accelerate claims resolution by reducing human touch. Specifically, the client intended to use AI to assess vehicle damage and estimate repair cost from the photos shared by the claimants.

To meet the stated objective, the client wanted to

  • Perform a discovery to identify the most impactful image analytics use cases that can deliver measurable business impact
  • Create a roadmap that outlines the execution approach, data requirements, and the required technology stack

Challenges

  • Lack of buy-in from the senior leadership team to invest and leverage AI and ML
  • It was unclear what aspects of repair costs can be ascertained from images

Solution Methodology 

  • Identified three primary drivers of auto repair costs
    • Damage category: 12 distinct categories identified, including dents, scratches, cracks, stains, broken and missing parts
    • Damage severity: created a rating scale to ascertain moderate to significant damage severity
    • Body style: Sedan, Pickup, SUV
  • Recommended high-impact image analytics use cases for execution by selecting those that are most addressable based on body style and damage category
  • Developed a blueprint that outlines AI & engineering skills and the conceptual technology architecture required to address the key challenges below
    • Creating high quality annotated data
    • Accounting for variations in image quality

Business Impact

  • The roadmap enabled the client’s analytics team to secure business buy-in for the use of AI/image analytics in claims resolution
Download Case Study

©2017 Tiger Analytics. All rights reserved.

Log in with your credentials

Forgot your details?