Components of Customer MDM Solutions
Data Loading and Data Integration
Data Profiling
- Perform data profiling, data quality assessment, anomaly detection, and metadata discovery.
- Utilize prebuilt analyses to examine individual attributes, including minimum, maximum, frequency distributions, and patterns.
- Identify values that occur frequently and detect outliers or exceptional values.
- Identify and exclude junk values, generating a cleaning list for data improvement.
- Access packaged processes for common quality tasks, such as handling incomplete data, resolving conflicts in duplicate records, merging rules, auditing, and more.
- Present profiling results graphically using various chart formats.
- Generate textual reports that highlight profiling results for easy understanding.
- Prebuilt graphical dashboards that display profiling results, including junk values, out-of-format PAN, suspicious DOBs, and more.
- Schedule the execution of profiling processes using built-in or third-party scheduling functionality.
- Access standard reports that provide comprehensive visibility into profiling results and data quality metrics.
- Perform efficient parsing operations to extract and manipulate data elements.
Data Cleansing and Standardisation
Matching and Clustering
- Proprietary algorithms (CLIP for Bulk, Prime 360° for Real-time) convert strings to numbers and determine attribute match extent.
- Robust facilities in batch and real-time modes for cleansing, matching, identifying, linking, and reconciling customer master data from diverse sources, facilitating the creation and maintenance of a comprehensive customer's golden record.
- Achieve high precision and recall in data matching.
- Perform matching on all defined attribute combinations to address data inadequacies and optimize recall.
- Extend clusters by associating them with user-determined properties.
- Conduct network analysis for deeper insights and connections.
- Ensure high-performance operations.
- Address data inconsistencies and nonuniform attribute availability.
- Support multi-threading for enhanced efficiency.
- Run all matching rules simultaneously.
- Utilize clustering to link records belonging to the same entity.
- Perform extensive linking to achieve comprehensive results.
- Employ undirected weighted graphs for advanced analysis.
- Support dual clustering with clusters based on MPC, but prioritize LPC clusters upon manual verification.
- Classify and grade matches as perfect, authentic, system, MPC, probable, suggestive, referral, or LPC, thereby emphasizing high precision.
Data Stewardship and Case Management
API and Integration Channels
Matching Rule Configuration and Survivorship Rule Building
- User-friendly interface for creating matching rules
- Support for multiple Matching Rule Profiles (MRP) with the flexibility to choose one before submitting a request. MRP consists of multiple rules with an 'OR' relation.
- Matching Rules allow for AND/OR operations between attributes.
- Option to treat an attribute as optional, matching if available, or considering it as a match even if it is 'NULL'.
- Flexibility to apply multi-value parameters for cross-referencing matching or matching specific types.
- Adjustable tolerance for each attribute's matching set, allowing approximate matching for attributes like DOB, Contact No, and Identifiers.
- Variation in matching tolerance can be set for different rules.
- Ability to search on complete data or subsets of data (Confinement).
- Confinement can be applied at the rule or MRP level to enforce all rules.
- Rules to assign preference to the most reliable sources.
- Dynamic confinement settings can be defined while building the rule or deferred to apply at runtime when the request is posted.
- Residual attributes can be designated, contributing to match confidence assessment without participating in the matching process.
- Assign weightages to attributes to calculate match scores effectively.
- Results can be classified and labeled into different categories based on business rules.
- Ability to grade match quality for each category.
- Rank results to prioritize the best matches at the top, with lower ranks indicating higher match quality.
- Log creation for rule creation activities.
- Intuitive interface for defining Survivorship rules.
- Attribute values can be determined based on Survivorship rules, considering factors such as source, timestamp (aging), latest values prevailing over older ones, longest values, maximum, minimum, average, etc.