Overview

Atto’s Categorisation Engine empowers you with deeper insights from transaction data, delivering accurate and automated transaction enrichment. This solution helps you understand how your customers manage their finances by classifying and categorizing every transaction.

Using advanced machine learning and big data, our engine analyzes transaction descriptions, assigns them to predefined categories, and predicts the merchant name for each transaction. This gives you a rich, comprehensive understanding of customer spending habits and income sources.

Categorisation, Classification & Merchant identification can be used in two ways via our APIs:

• Open Banking Journey: The end user connects their bank account via Open Banking. We retrieve and enrich the necessary data.
• Categorisation Engine: Customers feed their own transaction data into the API for processing, bypassing the Open Banking journey.

Challenges We Help Address

• Understanding Spending and Income Patterns

  Gain a clear view of how end users spend and receive funds across different categories and payment methods.

• Predicting Merchant Names Accurately

  Identify merchants from incomplete or ambiguous transaction data to improve the quality of insights.

• Enhancing Data Accuracy and Confidence

  Our model assigns confidence scores to each classification, ensuring high accuracy and reliable transaction enrichment.

• Reducing Manual Effort and Operational Costs

  Automate transaction enrichment processes and minimize the time spent on manual classification.

• Improving Customer Segmentation and Personalization

  Use enriched data to personalize offerings and enhance customer experience by understanding spending patterns in detail.

How It Works

Transaction Categorisation

Our engine assigns each transaction to one of 81 predefined categories, such as:

• Electronics
• Supermarket
• Salary

These categories provide a clear and comprehensive view of spending habits and income sources for each end user.

Transaction Classification

The Categorisation Engine also reveals how funds are spent or received by classifying the payment method for each transaction. It supports 11 transaction classes, such as:

• Point-of-Sale (POS)
• Cheque
• ATM

Merchant Identification

Categorisation is critical for understanding the type and scope of an individual’s spending. However, identifying the specific merchants with whom account holders do business provides a deeper level of insight.

Why Merchant Identification Matters:

• Tailored Financial Advice: By focusing on the merchants involved in transactions, financial institutions can offer personalized services and advice based on real spending behavior.
• Improved Customer Assistance: Understanding specific vendor relationships allows for more targeted customer support and product recommendations.
• Enhanced Behavior Analysis: Merchant data reveals detailed insights into banking habits and transactional behaviors, enabling better customer profiling and segmentation.

This merchant-centric approach offers a more sophisticated view of how funds are spent and helps financial institutions deliver tailored, data-driven services.

Confidence Scores

For categorisation & classification our model provides a confidence score (0.0-1.0) to indicate the reliability of the assigned category and classification. This ensures that the enriched data is highly accurate and trustworthy.

API Response Fields

Category

• id: Category id
• name: Category name (e.g Salary)
• confidence: Confidence score (0.0-1.0)

Class

• id: Classification id
• name: Classification name (e.g ATM)
• confidence: Confidence score (0.0-1.0)

Predicted Merchant Name

• predictedMerchantName: Predicted merchant name.

API Schema and Response

• Stored Data API (Get Transactions)
• Categorisation Engine

Use Cases

With Categorisation & Merchant Identification, you can:

• Understand Spending Habits: Identify where and how customers allocate their income.
• Detect and Analyze Income Sources: Gain insights into the types of income customers receive.
• Improve Lending Decisions: Use categorized and merchant-enriched data to assess a customer’s financial behaviour.
• Personalize Customer Experiences: Develop tailored financial products based on spending patterns and vendor relationships.
• Optimize Loyalty and Marketing Campaigns: Track merchant-specific activity to create targeted offers and improve engagement.

Key Features & Benefits

• Accurate and Automated Transaction Enrichment

  Reduce manual categorization and merchant identification efforts with our automated machine learning solution.

• Enhanced Customer Insights

  Access a deeper understanding of customer spending patterns and specific vendor relationships for more informed decision-making.

• Confidence Scores for Reliable Results

  Confidence scores help ensure the accuracy of each transaction’s assigned category, class, and merchant name.

• Improves Operational Efficiency

  Automate data processing to save time and resources while improving the quality of insights.

• Supports Better Decision-Making

  Clean, structured, and enriched transaction data forms a solid foundation for risk assessments, customer segmentation, and targeted marketing strategies.