With the release of DQC’s AI agents for data improvement, customers can now improve their data automatically, saving hours of manual effort each week and enabling more reliable, higher-quality data-driven work across systems.

Key capabilities

DQC’s AI agents can be used across a range of data-cleansing and enrichment tasks, including:

• Correcting and completing addresses using internal, external, and web sources (e.g., OpenCage)

• Identifying and merging duplicates, even across datasets and naming variants

• Applying industry classification standards such as ETIM, ECLASS, GS1

• Standardizing formats such as units, dates, and telephone numbers

• Enriching missing or incomplete records with contextual external data

• Executing complex operations using custom Python logic where needed

• Plus many additional functionalities that are either in private-beta mode, or company-specific (don’t hesitate to reach out to your DQC contact!)

These tasks can be configured individually or combined within DQC workflows. The AI agent operates based on users’ input and adapts to specific project requirements.

Configurable and context-aware

To improve accuracy and relevance, customers can provide the AI agent with:

• Supplementary background information (e.g. guidelines, documents, sample records)

• Access to internal systems (e.g., SharePoint) or external sources that may be useful

• Instructions or examples that define what good looks like

The agent adapts dynamically based on the available context and becomes more effective the more guidance is provided.

Full traceability and transparency

All outputs generated by the AI agent are fully documented to avoid potential hallucinations:

• Each action is explained in clear, easy-to-understand language

• Source systems and external references are listed

• Confidence intervals are included for every suggestion

• Output can be reviewed, verified, or programmatically approved

This ensures customers retain full control and visibility over automated changes to their data. It also ensures better data with AI plus human intelligence and control.