Elka,
I don't believe this is something you can do directly. A subset is not something that is directly callable. However, there are a number of other ways to accomplish this, depending on the volume of the data that is involved. If you were expecting a result set of 5k or less, you could probably create a clone data pool. To do this, you would create a subset extract definition with the tables you wanted included. In the extract definition, include the variable you want to use for the name field where you want to match 'Jon' or whatever other name interests you. Then, create a cloning data pool. When you set up the data pool, don't use any transformations. It will pull all rows that match the extract definition.
If you have a larger data set, you may be better off using a post-publish action and having Javelin orchestrate the subset for you.
Thanks,
Brian