AI-based database cleaning

The project

Most companies working with data will likely encounter some datasets which contain incorrect or missing records. This corrupted data can have a negative influence on your work. Cleaning a database can take up a lot of time, especially when workers have to manually check for the corrupted data to be able to change or delete it. The AI-based database cleaning tool is designed to help with this process in an automated way. The tool first analyzes all records, and then prompts the user with a validator for marking obsolete and missing records. A user will still have to take a look at individual records to conclude whether they are corrupted, and to change or delete them manually, however the time required for searching these records is cut down dramatically. The product allows one to export the list of records and put them next to the excel worksheet to change the corrupt records. Additionally, errors that have been reviewed by the user can be automatically repared in the original database by creating and executing a generated SQL script with fixes. This makes a huge decrease in the workload involving database cleaning. Our product is implemented as a general framework, which makes it possible to use it in various business environments.


The customer

Our costumer is Dr. Ramackers of LIACS. He was very involved in the process of our project. During our weekly meetings, he was always available for questions and feedback. Some weeks, mostly when the sprints started, we gave a demo of the implemented functionalities. His feedback was always very clear and he was open to any questions from the team. The product owner and scrum master were both in contact with the costumer. The product owner focused on presenting the demos and other product related materials. The scrum master kept the costumer up-to-date about the weekly meeting and other planning0related subjects. Communication during the project was very good, with quick response times and a professional atmosphere.

"Software Engineering was very helpful in getting to work with scrum"
The team

Our team dynamics were good, despite thefact that we had to learn a lot throughout the project This was for most of us the first experience with scrum, so this was something we needed to adapt too. Communication is something we worked on a lot. Because of the pandemic situation, we decided to do our weekly meetings online after a couple of meetings on campus. Meetings through MS Teams worked out just fine and our costumer was happy with it too. Keeping each other up to date on the work was a bit harder, also because of other subjects and their corresponding deadlines. But after a while we have gotten better. Our biggest obstacle was the loss of a team member, this was for us all difficult to deal with. However, after some advice from our costumer, we dealt with the situation and got through this obstacle. Over all, we have experienced a good cooperation within the team.


The technologies