Hi , I dont see a function toNumber in the record path syntax , so Im not sure how did you come up with this. It would be helpful next time if you provide the following information: 1- input format. 2- screenshot of the processor configuration causing the error.   As for your problem , the easiest and more efficient way - than splitting records- I can think of is  using the QueryRecrod processor. lets assume you have the following csv input:   id,date_time 1234,2024-11-24 19:43:17 5678,2024-11-24 01:10:10     You can pass the input to the QueryRecord Processor with the following config: The query above is added as a dynamic property which will expose new relationship with the property name that you can use to get the desired output. The query syntax is the following:   select id,TIMESTAMPADD(HOUR, -3,date_time) as date_time from flowfile     The trick for this to work is how you configure the CSV Reader and Writer to set the expectation on how to parse datetime fields in the reader\writer services: For the CSVReader, Make sure to set the following: CSVRecordSetWriter:   Output through Result relationship:   id,date_time 1234,2024-11-24 16:43:17 5678,2024-11-23 22:10:10     Hope that helps. If it does, please accept solution. Thanks   ... View more

Hi, Im unable to replicate the error . Can you provide more details about your flow with the processor configurations. here is what I tried and it worked:     GenerateFlowFile: EvaluateXPath: Output flowfile attributes:   ... View more