We had a long chat about the pilot projects that Capgemini is running for the government in relation to the G-cloud programme. I will be writing an analysis of this next week but thought I would get some of it out of my system.
The interview left me thinking that the next government, whoever it may be, could make big mistakes by targeting IT projects to meet cost cutting targets. If the wrong projects are canned the costs in the long term and loss could be huge.
Capgemini is one of several suppliers that are creating pilots for the government. These pilots are being done to educate the government. Although I am sure that the suppliers quite rightly hope to win business, Deacon says Capgemini is actually already doing this, so the government pilot is a learning curve.
Capgemini is building two full cloud infrastructures with one using Microsoft and the other the Cisco, EMC and VMware partnership. It is then building development platforms on top of these infrastructures. The final layer in the cloud will be software as a service.
I recently did an interview with Wipro's CTO I Vijay Kumar about what the supplier was cooking in the cloud.
He told me all about how by applying cloud computing to its core business of software development it had reduced the time taken to set up a project from weeks to minutes.
Deacon at Capgemini described to me some of the savings that the government could potentially make by using the cloud for its IT infrastructure, its software development and its application delivery.
And the snowball effect of time and money savings will provide free up the resources required to transform how the government goes about its business.
In relation to the infrastructure layer the costs related to ICT in government can probably be cut by 20% by moving to the cloud. This will cut costs associated with hardware, power and time for example.
But this is just the tip of the iceberg, although it is at the bottom.
When you get to the next layer, which is the development platform, things get interesting. You can have all applications or fragments of applications in a library. The IT department then pieces fragments together in the cloud to create a service for a specific department. Fragments of code might for example include a credit check or a citizen identification system.
So new code would not have to be written ten times over for new applications. Add to that the fact that all departments can share and you have massive savings. Then, like Wipro, the development process is streamlined with computing power available through the cloud on demand.
Then you have the third layer where the applications sit. All departments will have access to the final applications through the cloud. No need for lengthy procurement cycles.
But possibly the most interesting point that Deacon made is the fact that by saving money and freeing up resources the real benefit is to re-engineer government processes across departments. This flexibility could mean even bigger savings in the future.
"This will give the government the money and the time to develop shared services, which in the past have been to expensive."
Deacom says Capgemini is already using these cloud services in some government departments but expects a broader take up next year. He predicts full scale cloud computing in government in 2014, so just when the government will change again?