AI Claims Automation helped a general Insurance Giant to enable automated claims (damage assessment) in under 180 seconds.

The Background

Auto insurance is a policy purchased by vehicle owners to mitigate costs associated with getting into an auto accident. The challenges faced by traditional auto insurance claims system are :

  • More than 90% of processes are manual which results in high labor costs, more errors & poor experience.
  • Manual vehicle inspection takes time & prone to errors
  • Every claim cost huge losses to businesses
  • Time taking & slow assessment that takes 3 to 15 days.
  • Customer experience becomes poor and dicey
  • Legacy processes, rule engines, and silo systems increase inefficiencies, increase risk & fraud
  • The claims assessment is done without any learning from the past which increases the chances of fraud & error.
  • Lack of data insights barrier for their data-driven risk, profiling, and decisions

Key Requirements & Challenges

  • Simplified vehicle claims assessment process.
  • Increase customer trust by reducing fraud & errors.
  • Image & Video based real-time claim price estimation
  • Technology-enabled the detection and classification of damages.
  • Building customer credibility by accurate price calculation & fraud detection.
  • Automated invoice reading and mapping.

About Life Insurance Giants

  • >2 Million Motor Policies Issue Per year
  • 7000+ Branches PAN India
  • 10000+ Employees
  • 10+ Systems used for claims
  • Public Listed General Insurance Company
  • Approx. claims time 48 hours to 7 days

Before AI-Based Auto Claims

  • >70% Manual Process
  • >2 -7 Days Turnaround time
  • <20% Data-Driven Insights
  • Error-prone insights
  • No Scoring, no risks, no profiling
  • Legacy process and too many systems
  • Manual video or image-based inspection
  • 100% Surveyor dependent claims process
  • Bad customer experience
  • Increase in fraud cases

After AI-Based Auto Claims

  • Technology-enabled real-time vehicle damage detection and assessment
  • Easy calculation and classification of vehicle damage
  • Historical Data based matching
  • The manual intervention reduced by 75%
  • 100% Automated end to end process for policy issuance
  • Real-time verification to reduce errors & fraud
  • Digilocker based DL/RC Digitisation for real-time assessment
  • Development & Infrastructure costs were reduced by 85%
  • Turn around time reduced to 15 minutes with 80% accuracy
  • AI/ML-based guided solution for video capturing, image capturing
  • 360-degree video analysis to find instant damage & estimated price
  • No-touch and self claims assessment platform

MAQ’s Approach

MAQ Computer Services LLC has simplified the entire claims assessment process using image recognition, deep learning & neural network to increase customer trust as well as reduce the fraud, and turn-around time with fewer errors. Technology-enabled real-time auto/vehicle claims assessment to identify parts, damage detection, and intensity of the damage. Classifying the damages for repair, replacement, and painting with high, low, and medium severity. Calculate the price based on vendor, location, state, and damaged parts in real-time to get an assessment. Fraud detection to prevent misuse of images, make-n-model match & verification. Historical data-based pattern matching for fraud potential intelligence to reduce risk. Self-training image annotation engine to train new vehicles, images, and damages for improving accuracy & accurate claims assessment. The claims process now takes less than 15 mins.

Result of AI Based Motor /Auto Claims

  • Up to 300% Improvement in claims processing time
  • Reduce TAT up to 75%
  • Fraud image detection was reduced by 40%
  • No-touch process improvement by 70%
  • Up to 50% increase in customer happiness
  • Reduced per claims cost up to 40%
  • Automated claims issuance improvement by 20%
  • Improved predictive decisions up to 70%
  • Up to 25% increase in operation inefficiencies
0 0 votes
Article Rating
Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments