The Challenge

In a large public institution responsible for social security oversight, employees regularly needed access to company-related data for inspections, audits, or policy work. Typical questions included:

“Show all company registrations in the Ghent region from the last quarter.”

These requests were traditionally routed to the central data analysis team, who manually translated them into SQL queries. This process created bottlenecks, with employees often waiting days or even weeks for answers — slowing decision-making and increasing dependency on technical staff.

The Solution

To solve this, Common Sense AI, in collaboration with the organization, developed a text-to-SQL solution powered by generative AI.

The system allows users to ask natural language questions, which are automatically converted into SQL queries and executed on internal databases, giving non-technical users direct access to real-time data without needing to write code.

Key features included:

• Full local deployment to ensure data security

• A working demo based on open flight data to showcase the system’s real-time capabilities

• Tailored integration with internal systems to ensure reliability and scalability
  
  

The Result

The first tests with business users, including those without SQL knowledge, showed immediate and measurable benefits:

• Query response time reduced from weeks to minutes

• Significant reduction in workload for the data analysis team

• Improved satisfaction among internal users, thanks to increased autonomy and faster insights

This project marked a successful step toward democratizing data access within the organization, with future iterations planned to improve user experience and expand coverage across additional data domains.