Back
Key Takeaway
Building a natural language-based audit analysis environment for data exploration without SQL
We established a Text-to-SQL environment that allows structured data to be queried and analyzed in natural language, laying the foundation for audit work innovation where non-developers can directly verify scenarios and explore data.
Alcoholic Beverage (H Company)
Client :Alcoholic Beverage (H Company)
Industry :Manufacturing
Service Area :Data & AI
Applied Solution :AIR
1. Overview (Project Background)
This project was pursued with the goal of querying structured data in natural language,
and building an environment capable of automatically generating natural language responses based on the queried data.
In the existing audit system, SQL-based queries were essential,
which created limitations in data accessibility for non-developers and verification of diverse audit scenarios.
Accordingly, the main challenge was set to improve data access methods more intuitively
and simultaneously enhance both efficiency and analysis quality across audit operations.
2. Solution (Resolution Approach)
In this project, we introduced Text-to-SQL functionality to the audit system,
building an environment where diverse audit scenarios can be verified quickly and accurately through natural language queries.
Through this, non-developers can now explore data without directly writing SQL,
and with natural language-based query and response methods, rapid insight derivation and improved analytical capabilities have become possible.
As this autonomous data exploration experience accumulated,
the scope and frequency of data utilization naturally expanded.
Additionally, data accessibility and usability have been significantly improved,
creating an environment where anyone can easily query and analyze data,
and based on this, diverse audit scenarios can be repeatedly verified,
improving the overall efficiency of audit operations.
Meanwhile, as responses to simple query requests decreased,
and development personnel have secured a structure to focus on higher value-added tasks
such as data quality management and service enhancement.
3. Result (Achievements)
Through this project, we validated the potential of a natural language-based data query environment,
and for future user expansion and application to more diverse business and audit scenarios,
we are considering the following development directions.
Enhancing usability through feature advancement and user experience (UX) improvement
Expanding the scope of applicable audit and business scenarios
Building a stable operating environment through security system and infrastructure optimization
Through this, we have established a foundation for gradually expanding the natural language-based data utilization environment.
