Creactives data assistants support requisitioners that are procurement’s internal clients by understanding their purchasing needs as described in their own natural language.

Matcher and MG Prompt facilitate easy requisitioner discovery of the item(s) they need within existing master data or catalogs. If there are no matches, they properly categorize the new requisition. This helps procurement optimize processes and PO flows by minimizing incorrect categorizations that would otherwise lead to wasted time and money.

Matcher easily and intuitively solves the first problem requisitioners often face when they create a purchase requisition (PR): the difficulty of finding the items they need within the material master data or internal catalogs and frame contracts. The difference between how the requisitioner describes the item he needs and how this item has been described within the ERP data often creates a barrier for users. Matcher, fully integrated in the PR transaction, understands requisitioners’ natural language and proposes the 10 most similar items. In 95% of cases, one of the proposed items is the correct one.

MG Prompt supports requisitioners when no matches are found and they need to categorize using a free text order line. Again, MG Prompt, fully integrated in the PR transaction, understands the natural language of the requisitioner and proposes the 5 most comparable categories. In 95% of cases, one of the proposed categories is the correct one.

In combination, Matcher and MG Prompt ensure accurate data from the very beginning of any PR and make the procurement processes smooth and effective.

Last but not least, UNSPSC Prompt is the prompter to support UNSPSC categorization of any item by understanding more than 25 different languages.

Data assistant at a glance:

  • Ex Ante Approach: Creactives assists in the creation of the right data from the beginning; providing reliable data to platforms such as QlikView, Tableau or Power BI.
  • Full Integration with ERPs: Reduce free text requisitions and orders;
  • Continuous Improvement: Every user interaction improves the machine learning model;
  • Natural, Native Multi-Language: Over 25 languages including Chinese, Japanese, and Korean;
  • «Easy to adopt» Solution: No training is required to integrate the tool in users’ daily activities.

Related Case studies


Contact Us