Overview
Customers’ account history holds valuable information. Augmenting raw data with spending categories allows better understatement of customer habits. It allows customers to manage their personal finances in a straightforward manner, analyze it and take action. Financial institutions can use this additional information in internal processes, customer segmentation and use it as a main value driver.
Websensa PFM uses advanced AI-based and Rule-based classifiers to label each user transaction with appropriate category,. Classification Engine uses custom, multi level category tree that will be best suited for the Bank and it’s customers needs.
Advanced Architecture:
​
-
Multi-Level Architecture leveraging
-
Artificial Intelligence and Rule-Based Classification,
-
Automatic adaptation to user behavior,
-
Customizable Category Tree
Characteristic
High Accuracy:
95%
transaction labeled correctly
High Efficiency:
150 transaction per second per core
High Scalability:
Real-time transaction categorizastion
Key features for customers
Limits, Alerts & notifications
Provides a mechanism that identifies and detects events, which trigger notifications to customers when certain events occur, e.g. a limit is reached or a new transaction is received.
Budgeting
Set targets connected with specific categories, subcategories or merchants. This allows customers understand how they spend their money compared to a plan and helps to make decisions to change their financial behavior to fulfill or change targets
Cash flow prediction
Identify and manage cyclical payments and incomes. Give customers a comparison between current and historical flow. Predict future transactions and payments.
Financial analysis
Create custom reports and views of transactions based on categories, subcategories or merchants. See aggregations of spendings per week, month or year.
Spending/income categorization
Automatically assign each transaction to a category and subcategory. That gives customers an insightful overview of their spendings an incomes.
Key features for financial institutions
Client segmentation & risk assessment
Builds customer representation based on multiple data, like: static features (demographic data, geo location, devices), owned products, cash flow, spendings categories distribution. Use it in recommendations, prediction services, credit scoring or fraud detection. This works in real-time, based on current user activity.
Customers data augmentation
Enhanced your data that can be used for multiple products and helps with customers behaviour analysis.
How it works
Integration
REST API - solution can be easily integrated into your existing infrastructure using API. Used when near real time is required. Transactions are classified as soon as they become available for the engine.
​
Batch mode - used for more efficient classification of large amounts of historical data.
​
Custom integration - PFM is open for other ways of integration. Queue systems are supported as well as direct database access.