Our unique approach to System modernization has been designed to minimize errors and data loss:
Many legacy systems suffer from poor data structure. The information had been intensively duplicated over the years, resulting in some tricky situations where the same piece of data was contained in several tables. There was no referential integrity, and the concept of primary and foreign keys wasn't applied uniformly across the database.
Our first objective is to modernize the database to create a clean and easy-to-use information system. We design a relational database applying the normalization. To facilitate the reuse of the data across the organization, we split the database into two different schemas:
One was dedicated to corporate data, reusable in many different applications.
One was holding data specific to the legacy application we had to modernize.
Then we populated the new database from the legacy database using the Extract Load and Transform (ETL) strategy. So we ended up with an information system capable of providing data for the client's system, but also with a more generic layer dedicated to corporate data.