I will answer broadly regarding Xcalar's features and refer you internally regarding facets that touch on your engagement with Xcalar's field team.
You asked a number of questions:
1) If you want to run your batch dataflow by cron job, you can accomplish that by using Xcalar's built-in scheduler to run your dataflow according to your cron scheduling expression (like: * * * * 23 *). I understand you have business requirements to use a CI/CD tool to schedule your dataflows; you will need to sync up with your team on the details of the pilot for your company.
2) Regarding customizing the input file path: instead of customizing, I'd recommend using a regular expression or configure import by directory using a wildcard and recursive search to point at the input files.
3) Regarding custom schema: I understand that your custom import process will use a configuration file containing, among other things, delimiter and schema for input files. Again, you should sync up with your team regarding the pilot.
4) Regarding output path: You can parameterize a dataflow's output to use a custom directory name including the parameter N, representing the dataflow execution number. That will insure that your output path is always unique.
Xcalar is committed to getting you set up using one dataflow; that way you will have the consistent inputs and outputs necessary for your company's use.
Do you have any additional questions about this?