Automatic transaction categorisation in Embat

TellMe, Embat’s AI-based treasury analyst, works in the background to keep financial information as accurate, consistent and up to date as possible. One of its key functions is automatic transaction categorisation, an essential process to save time, reduce errors and ensure reliable reporting.

What is automatic categorisation?

Automatic categorisation assigns an appropriate category to each bank transaction (payroll, taxes, fees, income, etc.) based on historical patterns, rules defined in the platform, and Embat’s new AI model.

Its purpose is to:

• avoid repetitive manual work,
• improve the quality of bank data,
• simplify the creation of rules,
• provide more accurate dashboards,
• support more reliable cash-flow projections.

Each properly categorised transaction helps build a clearer, more up-to-date and actionable cashflow.

How TellMe works in Silent mode

TellMe operates continuously through its Silent mode, running background analyses and executions to process, enrich and categorise data automatically without requiring user approval. Even so, the user retains full control, as any applied categorisation can be reviewed and modified from the Transactions module.

A new model with 95% coverage

Automatic categorisation uses TellMe’s new AI model, which reaches up to 95% coverage.

Coverage refers to the percentage of transactions that the system can categorise automatically with a high level of confidence.

In practice:

• the vast majority of transactions are classified without human intervention,
• the number of “uncategorised” movements decreases significantly,
• finance teams spend less time on manual tasks and more on analysis and decision-making.

How does TellMe decide the category of a transaction?

TellMe combines several criteria to determine the most suitable category. Each categorisation shows its origin to ensure full transparency.

The possible criteria are:

• Category suggested by TellMe: The transaction has been automatically categorised by the AI model based on learned patterns and rules.
• Based on your history: The category normally applied to transactions with a similar description is assigned, based on the user’s or company’s historical behaviour.
• Pattern detected: TellMe detects that other transactions from the same bank with similar descriptions are usually classified under that category.
• Accounting inheritance: The category is inherited from the accounting entry with which the transaction was recorded in the ERP.
• Debt management: The category is applied according to the settings defined in the loan amortisation schedule.

This combined approach allows TellMe to learn and improve progressively.

Prioritisation: rules first, TellMe afterwards

If a rule exists in the Automation module, it will always take priority over the AI.

The flow is:

1. Rules: if a rule matches the transaction, it is applied automatically.
2. TellMe: if no applicable rule exists, TellMe categorises the transaction.
3. Manual review: if neither system can determine the category, the transaction remains “uncategorised” for user review.

Over time, thanks to the model’s learning, the number of uncategorised transactions decreases significantly.

Where to view and edit categories

All categorisation information is managed from the Transactions module. From there, you can:

• see whether a transaction has been categorised by a rule or by TellMe and review the criterion used for that categorisation,
• easily identify movements requiring manual review and edit the category at any time.

This ensures that the classification always reflects the company’s financial activity accurately.

Benefits for treasury management

Automatic categorisation provides direct benefits:

• Time savings: eliminates repetitive manual work by classifying thousands of monthly transactions.
• Higher data quality: improves the consistency and reliability of banking information.
• More effective rules: a clean dataset allows for simpler, more robust automations.
• Stronger reporting: dashboards and analysis built on better-categorised data.
• More accurate forecasting: cash-flow projections based on a coherent and up-to-date historical series.

TellMe ensures a clearer, more structured financial foundation that is ready for critical decision-making.