I had been thinking of building something like this, and was quite excited to find it. My team really needs this. However, after 30 minutes of evaluation, I must conclude this product is very Beta, I am not sure why this is even being sold yet:
1. The creation of a project went fine, connection to DB and SVN tested OK, until it came time to commit, when I got an error "commit failed access to svn act forbidden visualsvn". Some googling shows this was a problem with the case of the user name or path. Indeed, whoami shows my name as John.Smith and VisualSvn server as john.smith.
So I install a proxy (that does not support NTLM) between my computer and SVN, get prompted for a login, login as john.smith and everything works.
2. After everything is committed, every time I check, there are 2 tables that show as needing to be updated. I try to commit them, and I get an error 'Nothing to commit', which is indeed accurate because the diff shows no differences. These are not external tables.
3. So now I decide to try a more permanent solution (a permanent proxy rather than one running on my machine) and want to change the address of the SVN server. But there is no place to edit the settings. Not sure if it offended the sensibilities of a graphics designer to have a project settings screen, but you need one. I could have specified the user name I want to use to connect to svn (rather than have it inferred from the OS, and that would have solved problem 1) and I could have edited my project. What if I had 25 projects and the SVN server was decommissioned and moved to another machine? would I have to delete and recreate 25 projects.
4. The way the system is designed seems wrong to me. When I check in, I would like to see only the objects I modified. I do not want to accidentally check in another dev's changes, because that would then mess up the review process with Crucible. I know this program works without modifying the database, but if you installed a system trigger on the database, you could detect and record all object changes, get the username from v$session, and present the users with only the objects they modified. (And the commit could also be done directly from the database to the SVN server, with a Java stored procedure and SvnKit), without going to the client (I work remote, so all this network traffic only slows things down).
I am always too early, then late, evaluating products (Jira, Confluence sucked when I first tried them, now they are great). I will check back in a couple years and see if things have improved.