Sinclair Schuller over at Apprenda is someone I’ve been following now for a number of years. He’s obviously a vendor, so sometimes riffs off things with his own vendor-centric perspective, but that said he’s keen to push the thinking on SaaS generally, regardless of whether there is a “fit” for Apprenda in what he’s discussing.
Recently a post of Schuller’s caught my eye, in it he was discussing the ramifications of SaaS for government software needs – specifically the potential of private clouds and semi-generic apps to add real value, and avoid many of the concerns, around cloud computing.
In his post Schuller avoids much discussion of traditional public SaaS products (the example he gives is the grandaddy of SaaS, salesforce.com) quite rightly pointing out that the biggest concerns around SaaS in the public sphere is the security concern. Despite the best efforts of the evangelists among us, this fear is going to take awhile to go away, and we should look for other ways to introduce cloud computing and SaaS concepts within the public sector that don’t necessarily rely on storage outside of Government hands.
Instead Schuller suggests the leveraging of SaaS architecture to create what he calls “home brew applications” or customised SaaS applications on private clouds. The example Schuller gives is of one state producing an application to handle, for example, parking violations. In all practicality the next door state is creating their own application with quite possibly the exact same functionality. Now imagine some “vanilla” applications created centrally and hosted on private clouds.
Now imagine this central agency that creates the applications being able to tender for application provision to Government agencies. These applications, hosted as they would be on private Government clouds, would meet the security requirements of the particular customer agencies, but they would also leverage the fact that these vanilla applications could serve many customers across agencies, states and possibly functional areas.
While I’d never be a proponent of yet another Government agency, the fact is that all these individual departments quite likely pay millions of dollars to independent contractors who are, to put it bluntly, rehashing an existing code base with every new contract they get. The time and money that could be saved by moving this function away from a distributed, private sector pool, and onto a centralised, transparent and synergised agency is massive.
It’s an idea that has legs – it fulfils some of the efficiency related promise of Government 2.0, it avoid some of the potentially rort type contracts that seem to occur with all too much frequency, and as a side benefit it opens up the potential of data sharing and the benefits that go along with that. It’s a nice combination of public sector ring-fencing, with inter-agency code sharing… an idea whose time has come?