SPIKE: Design approach to remove post-processing step when Instance is created

Currently incoming MARC Bibs is stored in SRS, then Instance is mapped in mod-inventory and saved in mod-inventory-storage, where Instance UUID and HRID are generated. SRS MARC Bib is then updated with Instance UUID and HRID - we refer to it as to a post-processing step. 

Design the approach in which data flow happens in one direction - UUID and HRID for the Instance should be generated earlier in the processing so that SRS MARC Bib could be saved with appropriated ids in the first place