An administrative assistant regularly needed to query the ECM/EIM system to retrieve specific documents: old contracts, deliberations related to a department, HR files, or public procurement records. However, multi-criteria searches required strong command of filters and syntax, which slowed down information access and sometimes required assistance from a subject matter expert.
DINA simplifies this process by enabling natural language search in applications:
- The user expresses themselves simply, as they would verbally: “I’m looking for supplier contracts signed before 2021 for the purchasing department”
- DINA understands the query, identifies implicit criteria (date, document type, relevant department)
- It transforms it into a structured search in the ECM or any other business application (complex form, Boolean filter, proprietary syntax…)
- The result displays immediately, without manual configuration
Observed benefits
- Accelerated access to relevant documents or data, without technical knowledge required
- Increased autonomy for non-expert users
- Fewer requests to support teams or tool administrators
- Better utilization of data already present in business systems
DINA integrates a semantic parsing engine and natural language query translation capability, able to interpret user intentions and dynamically convert them into structured queries compatible with ECM, EIM, or other business application syntaxes (Boolean queries, parameterized forms, proprietary APIs). Through a configurable semantic abstraction layer, DINA contextualizes searches (document types, dates, departments, authors…) and queries existing systems without modifying their structure. Integration via API or native connectors ensures rapid, secure deployment compatible with existing access rights