Logic_Ca_The – Making Statistical Data Validation Faster and More Accessible

Inspiration

I am a blind statistician working at the General Statistics Office of Vietnam. Every day, I work with large statistical surveys containing thousands or even tens of thousands of records. For sighted users, reviewing such datasets is already difficult. For someone using a screen reader, it can take many times longer.

Because of this, I have always looked for ways to automate repetitive tasks and make my work more efficient. My goal was not only to save time for myself but also to create a tool that could help my colleagues work more efficiently.

How I Built the Project

I have no programming background. I built this project entirely by describing my ideas and requirements to ChatGPT.

Instead of writing code myself, I explained every feature step by step, asked ChatGPT to generate the code, tested the software, identified problems, redesigned parts of the interface and workflow, and repeated this process many times until the application behaved exactly as I wanted.

This experience showed me that AI can empower people without programming knowledge to create practical software that solves real-world problems.

The Project

Logic_Ca_The is an accessible desktop application for validating statistical survey data.

The application imports survey data from Excel files and reads validation rules from another Excel file. This design allows the validation rules to be updated without modifying the software itself.

After validation, the application exports all detected logical errors into Excel files organized by reporting unit. It also provides tools to filter specific error types into separate Excel files, generate error summaries, and produce preliminary statistical summaries.

Excel is used for both input and output because it is the standard tool used throughout my organization, making the software easy for everyone to adopt.

The interface is designed to work well with screen readers while remaining simple and intuitive for sighted users.

Challenges

The biggest challenge was building software without knowing how to program. Every improvement required careful communication with ChatGPT, continuous testing, and many iterations before the software worked correctly.

Another challenge was ensuring that every feature was accessible with a screen reader while remaining easy for sighted users. Accessibility was considered throughout the design process, not added afterward.

What I Learned

This project completely changed my understanding of software development. I learned that programming is not only about writing code—it is about understanding a problem, designing a solution, testing it, and continuously improving it.

Most importantly, I learned that AI can help people with disabilities become software creators instead of only software users. With ChatGPT, I was able to transform ideas into a practical tool that saves time, improves data quality, and makes statistical work more accessible.

Built With

• python