Build an application where Mongo Bulk Write using Hadoop Map Reduce

This are my steps to include: I have to include Json file from S3. File contains uid,ag,gn,yob,scrtp attribute. I have to make key as uid and value as ag,gn,yob,scrtp value and that is included in HashMap collection. Reducer will reduce it according to key,hash value can also be calculated based on key as I have 5 database in MongoDB. What I am expecting result to be write into 5 Mongo database according to hash value and write should be bulk write, how to achieve this as above steps? Need a person who can do this in Hadoop Map Reduce in Java?