The introductory video and use cases are interesting from an ideological standpoint (which I broadly support but won't comment on here). Note also members of the CAMBIA team are authors on the recent paper (Jan 2018) "Mapping the global influence of published research on industry and innovation" (PMID 29319684) and it turns out another blogger has already written a usefully detailed review. To be honest, I had started on my post before seeing this and consequently nearly abandoned this one.
However, I decided to continue with a look through my personal utility lens (pun intended 🙂 and pick out a few complementary aspects, not to mention a few quirks. I started with an inventor search (of yours truly natch') which gave 97 hits but applying the select "Group by Simple Families = (On )" filtered this down to the 30 indicated below.
This worked pretty well in graphically tracking what were then SB (GSK after 2000) novel sequence filings but I note a) there is some "family inflation" since the non-redundant list I compiled some years ago was 22 (SciFinder makes this 23) but not 30 (e.g. CA, JPs as well doubles for Afc1 and Rfc1) b) 26/30 had extracted sequences (which should be all of them since I put these sequences in!) c) they had a problem pulling text for the EPs and d) while I'm not seriously going to quibble over "Aspi" instead of "Asp1" it does indicate OCR noise from somewhere.
(JFTR I no longer update my published patent collection because not only are they now rather stale beer but also the 11 granted applications are now well on their way to expiring! At the time EST mining, cloning and filing "genes as currency" was a lorra fun while it lasted, see PMID 8524376 , PMID 10845246 and this retrospectroscopic post)
Following through examples shows the LENS emphasis on patent sequence indexing (indeed they do some of this for the NCBI), as can be seen for Human AFC1 below.