Examples from prototyping work

The examples below are what we hope to achieve through our prototype version of the improved process.

Using an LLM for text summarization

  • Allows for faster reading by the claims adjuster:

    test image

Using an LLM for information extraction

  • Extract key pieces of information for better population of database:

    info extraction

Using an LLM for sentiment analysis

  • Detect tone of text, and potentially act on it:

    proto sentiment analysis

Using image recognition to frame vehicle(s) in pictures

  • Analyse images provided by customer:

    proto car recog

Using image recognition to detect damage

  • Assessment of damage based on picture:

    proto accident grading

Web Application to review/process claims

  • Have an application that ties in these tools together and enables users to process the incoming claims more efficiently:

    proto claims processing app