A leading global specialty provider of property and casualty insurance and reinsurance in the UK received business proposals and statements in an unstructured, semi-structured, and structured format. The different formats included emails and attachments of type word, PDF, excel, scanned images or as email body which prolonged the submission cycle, ability to diligently handle submissions and total business written. The company asked AdfarTech to study the existing landscape, identify opportunities and bring operational efficiencies to Underwriters.
AdfarTech developed an AI data ingestion system to overcome the manual pre-bind process by automatically abstracting all applicable data fields from business proposals. The implementation was based on its proprietary AI solution accelerator leveraging NLP and ML techniques. The extraction was presented with a recommendation (decline, non-binding, etc.) on inquiry status to the Underwriter with the ability to override. The system was integrated with Office 365 to collect and process each email submission through an AI engine and display the corresponding values on the visual interface. Key highlights of the solution:
- Automated near real-time extraction of required elements from source email and attachments with lightweight workflow, business rules, and visual interface.
- Accelerated implementation leveraging AdfarTech proprietary solution accelerators using Machine Learning and advanced NLP based algorithms like BERT etc
- Automatic retraining and self-learning of NLP and ML models
- AI-based recommendation to support decision making.
- Risk model calculation integration to the policy administration system
- Automated verification of insurance documents helping Insurer to achieve desired ‘Speed to Market’ and improve the operational efficiencies.
- More than 95% accuracy on extraction helps in larger straight-through processing.
- Customizable and Re-learning helps in continuously enhance the system.
- Overall >60% improvement in productivity as users can spend more time on quality work.