Introduction
Typical digital transformation and legacy modernization programs in an enterprise have issues with business processes and rules hidden in legacy systems. In these enterprises, analysts do not document business processes and rules in detail. The goal of the transformation initiatives is to extract the business processes, business rules, business policies, and compliance rules. Inability to extract them leads to problems in releases and business continuity. The complexity in the software code is the key factor of extraction. There is no correlation between software code and business language. The goal is to identify the components, rules, different conditions, and relate them to business language. During this process of identification, you can add the key data entities which are associated with business rules and conditions to the documentation.
Transformation initiatives include new data model development and relating it to the business logic and rules.
Insurance Domain
Business Analysts conduct interviews as part of the rules extraction initiative before processing the code. They gather the inputs and outputs of the applications during the interview. They analyze the documentation related to business processes and rules before the extraction process from the code. They functionally decompose the system into different system modules and components. The extraction process blended with knowledge from documentation and interviews will help analyze the system for business rules extraction. Domain experts can write rules after gathering them in interviews. Developers and architects can document rules after analyzing the code, batch jobs, database schemas, and technical documentation. A software tool that can extract the rules and logic will help the developers and architects to correlate their observations with extraction output. Domain experts can provide business context and applications to be mined. The mined and extracted rules need to be persisted and transformed for real business processing. Business semantics association with rules will help the rules developer and analyst. I remember getting involved in one of the auto insurance online website which involved the extraction of business rules from the legacy mainframe underwriting system. For auto insurance, each state has different rules and forms with different fields as the driving speeds and road types are different in U.S.A.
Auto Insurance Rules
Let us look at enterprise business rules specific to insurance enterprise. Different rules related to the product, underwriting, document, process, interface, and policies are gathered during extraction. The data entities related to business rules are identified and their detailed definitions are gathered. Business analysts perform functional decomposition to extract functions and sub-functions across the system modules. During the code analysis, developers extract business logic filtering the operations, and gathering the logic blocks. They store this information after extraction. Unstructured information related to underwriting, claims adjustment, claims survey, social media interaction, and third-party witness is also collated during the extraction. Unstructured information with geospatial information, location data, traffic patterns, weather events, health records, historical profile data, aerial, and satellite images help in identifying the digital trace of the customer in time. You need to ensure the customer profile has age, address, call logs, online history, and location data. You can create mashups of customer behavior with demographics, preferences, and historical data. Enterprise can benefit by getting customer insights related to the following:
- Risk Profile
- New Exposures
- Pricing Strategies
- Underwriting decisions
- Loss prevention Strategies
- Merit Rating
Auto Insurance providers are coming up with new offerings based on your driving behaviors. New business models are emerging with new offerings. The usage-based business model helps in associating the insurance premium prices with driving behavior. You can capture driving performance profile attributes which are listed below:
- Speeds
- Acceleration
- Hard Braking
- Weather patterns
- Traffic Patterns
- Pathway Hazards
- Routes
- Risks
Auto Insurance
After capturing the driver’s profile and behaviors, you can provide real-time feedback and present with alerts about first notice of loss. Historical behavioral data can help the insurance providers to provide a customer protection blanket. Insurance providers can process the claims seamlessly with customer’s historical data. Decision Support capabilities of the provider are enhanced by automating the following processes:
- Claims Approval
- Claims Assignment
- Regulations Accommodation
- Claims Scoring
- Fraud Detection
- Payment and Settlement
Fraud Alerts
Business rules extraction and persistence help when the fraud detection techniques change. The change of the rules is documented and persisted as rule package versions. The rules go through the cycles of updates, validation, and storage based on changes in fraud rules, new policy updates, and changes in economic conditions. New techniques based on AI, NLP, and Blockchain are evolving to accelerate the business rules extraction from legacy and validate them for new business process execution.