Government SaaS Grids – Reinventing Application Solution Delivery
The Canadian Federal Government recently announced plans to create a single IT department ‘Shared Services’ that will save $100-200m a year through consolidation-driven efficiencies. As highlighted in this fact sheet there is lots of ‘low hanging fruit’ for cost savings due to a spread of many different email systems, networks and data-centres.
This expensive complexity doesn’t stop there, it also continues into their enterprise application estate.
Like most large organizations Governments are organized hierarchically, with different departments for each of their main functions, like Education, Welfare, Tourism and so on, and the head of each department, the Deputy Minister, is directly responsible for the privacy and security controls of the information they process.
This means that typically each operates as a ringfenced ‘fiefdom’, because the simplest and most absolute method of achieving this information security is simply to purchase and operate your own IT estate: You buy your own servers and business applications and keep them entirely separate from any others. Yes this does provide the means for the Minister to ensure their data privacy compliance, but it comes at a very high price – For the taxpayer.
This provides a local solution to their own needs but creates two key global issues: 1) A huge amount of under-utilized IT server hardware, meaning highly inefficient use of taxpayers monies, and 2) a lack of ‘joined up working’ across agencies, resulting in poor customer service for citizens.
Furthermore there’s a third and headline critical problem : slow application solution delivery.
This refers to the end-to-end process of procurement and deployment that an IT team undertakes to deliver technology to meet end-user requirements.
Due to these departmental stovepipes typically this process features purchasing an entire ‘COTS’ (Commercial Off-the-Shelf) application, which is dedicated entirely to their needs. A purpose-built case management application for social welfare claims for example, from vendors like SAP or Oracle, or custom developed software, such as a payment processing app being deployed to multiple branch locations.
In the same fashion as the multitude of networks and email systems described above, this approach drives another layer of duplicated IT, the operational elements built into each of these applications, such as identity authentication and user interface, as well as provisioning the app and service levels. The challenge, and opportunity, is that a large enterprise has a multitude of other applications that each repeat this same common layer of software functions. Staff have to repeatedly log in to multiple applications and become familiar with each different one every time, and critically citizen data is equally repeated across all of them.
Ultimately this causes ‘innovation gridlock’ – The effect where the application estate has grown to become such a morass that the bulk of the IT budget is consumed in simply ‘keeping the lights on’, there’s little left for new innovations. Furthermore since a COTS approach is taken for these new projects this means application solution delivery is based on trying to absorb huge monolith systems each time, a process that can take years to finally achieve if at all.
The drive towards the future that Cloud computing represents is the offer of more agile approaches for these scenarios – The ability to implement lots of smaller projects much faster, and thus deliver a continual stream of value to end-users.
Shared Services Architecture – Private SaaS Cloud
This new paradigm can be achieved through leveraging all layers of the Cloud wave, and the power of the Shared Services model is demonstrated through the fact it’s also the same principle, in software terms, that can address these issues.
SaaS (Software as a Service) is an implementation of this approach, where one common function, like a CRM application from Salesforce.com, is made available to many customers from a single platform.
With Private Cloud referring to the internalization of Cloud Computing, it’s key to note that this therefore also includes this SaaS aspect, as well as IaaS. To date many enterprises have widely adopted lots of this infrastructure-level virtualization, which can achieve one level of shared service consolidation, but as the diagram highlights this offers the least tangible benefit for users.
In contrast not only does SaaS provide the one-to-many scalability but it also facilitates a modular customization approach, meaning this does not come at the price that every one is forced to use exactly the same app, indeed through their apps store every Salesforce.com customer can configure a unique mix of functionality n a modular, plug-n-play manner. Replicating this effect internally would enable application solutions to be delivered to end-users equally quickly and effectively, maximizing their perceived value.
In their white papers Apprenda describes how SaaS is not the same as virtualization but instead highly complimentary,it builds on top and applies the same shared services approach to enterprise software.
Their SaaSGrid technology makes this possible, it automatically introduces aspects such as full stack multi-tenancy, drop-in identity federation, infrastructure independent deployment and runtime execution, and customer facing provisioning workflows to new and existing application assets.
In short it provides the tools to re-engineer ISV and custom software to operate in a SaaS mode.
Government Community Cloud
By combining these tools with the relevant government standards, like those for identity federation, then governments can employ these tools to create ‘Government SaaS Grids’. These can be a core architecture component of building Government Community Clouds – Learn more in our upcoming webinar.
- » SAP embraces Microsoft for stronger preferred cloud partnership
- » How companies can tell good cloud sprawl from bad: A guide
- » How to create a cloud centre of excellence: A guide
- » How to make your SaaS startup stand out from the crowded cloud landscape
- » AWS, Azure or Google: Do the differences between cloud providers really matter?