@GABILOBATO, I have run into similar situations in the past trying to compare database data types to “seemingly” same data types from other systems or databases. Differing database systems treat numbers and dates differently and the database reads result in different JSON outputs. The SnapLogic pipeline validation output may not catch the subtle differences, so it’s hard to catch. (it’s rare, but it has happened)
Therefore, one thing that may help is to capture the database outputs directly to a JSON file and then inspect the output visually in the file. You may find subtle differences in the numerical and date output and may need to use a parseFloat(), Date.parse(), or similar methods to synchronize the types.