Ask your production data questions instead of building queries.
vCHAT is intended to simplify natural-language access to VIVALDI data and complement vDASH with flexible ad hoc questions: read-only at first, later possibly with controlled and explicitly confirmed actions.
Planning note: vCHAT is currently being planned, with completion targeted for Q4 2026. The target date reflects the current planning status. Write or execute functions are explicitly not part of the current product scope.
Natural language instead of database queries
vCHAT is intended to let users ask specific questions of the VIVALDI database in natural language, with large language models acting as a language interface to existing production, quality and traceability information. While vDASH is suited to regularly required KPIs and visualisations, vCHAT is intended to make freely worded, exploratory questions possible, without writing your own database query.
For reliable productive use, permitted data areas, roles and permissions, the traceability of the data used and the handling of unclear questions must be clearly defined. Write or execute functions, such as blocking a part via a confirmed chat command, are conceivable for later expansion stages but are explicitly not part of today's read-only scope.
Example use cases
- ‘Which defect class has increased the most for this item over the last four weeks?’
- ‘On which machines were the parts produced in which defect class X was found?’
- ‘Which batch or part group is affected by a specific quality anomaly?’
What vCHAT is intended to make possible
- Lower entry barrier to production and quality data
- Faster ad hoc analyses without manual database queries
- Supporting plant management, quality and production in finding information
- Using the traceability and quality data already built up as a source of knowledge
Interested in vCHAT?
Get in touch: we will let you know as soon as vCHAT is available for your use case.