Spend Docs
Datasets

Datasets Overview

A Dataset in SpendCraft is a clean, segmented, and actionable collection of spend transactions. It represents the "prepared" state of your data, ready for the core analytical functions of the platform.

🏗️ From Raw Data to Actionable Insights

Datasets are created by taking raw financial records from your Sources and applying specific filters—such as time periods, currencies, or business units—to isolate a specific scope (e.g., "All Indirect Spend for the US region in FY2025").

🛠️ The Role of a Dataset

The Dataset serves as the single, standardized input required for the platform's most critical processing actions:

  1. Vendor Normalization Automatically cleaning and consolidating inconsistent vendor names (e.g., merging "AWS," "Amazon Web Svcs," and "AMZN" into a single record).

  2. AI Classification Feeding data into a trained AI Agent to assign precise categories from your Taxonomy to every transaction line.

💡 Why use Datasets?

By segmenting data into targeted Datasets, finance and procurement teams can:

  • Improve Accuracy: Work on specific data segments for higher AI precision.
  • Granular Reporting: Isolate regional or departmental spend for deeper insights.
  • Standardization: Ensure all raw source data is mapped to a uniform format before analysis begins.

On this page