Imagine this: you pay your gas bill using a credit card and the application automatically fills in the form for you by pulling in data from the Post Office, a credit reference agency, the credit card company, the energy company, and the Home Office, all of whom know something about you that helps you complete the transaction successfully, and which none of them shares.
This is the personal vision of Marc Silvester, senior vice-president for research and global CTO of Fujitsu, which is celebrating its 75th anniversary this year. Silvester calls the idea "My persona".
"It is about pulling information together that is freely available, but impenetrable," he says.
It may be Silvester's private vision, but it is also driving some advanced R&D in identity, security and transaction processing at Fujitsu.
Silvester says R&D at the company has changed since the days of ICL. Rather than betting the company on ground-breaking technology, everything these days is associated with a customer, even if the pay-off is five years away. As a result, little improvements can yield big sales, he says.
This approach is very Japanese, the essence of kaizen. But it is not stopping the UK from leading in key research areas.
"Cloud computing today is about virtualised machines and processes, tomorrow it will be virtualised transactions," Silvester says.
Finding the path is Fujitsu's Hayes Park laboratory near Heathrow. Formerly the centre of the firm's grid computing research, it is now home to work on information identity, security, transaction design, mobile and "quantum networking", all of which are relevant to Silvester's "My persona".
Programming the coming transaction-oriented systems will require new techniques - something Fujitsu is working on. But there are already some early examples of what Silvester believes will become the ruling paradigm. One is the Government Gateway, the portal through which UK citizens increasingly transact with government. The other is Apple's App Store.
The key is to free data from its dependence on proprietary processing formats and architectures, he says. Data will have to carry its own identity attributes, security and processing instructions with it as it moves around a networked world. This will enable any system anywhere to process it correctly and to send the results back to the right place.
This model will increase hugely the traffic in the network, because of all the metadata that must accompany the data items. It will take new ways of getting them into and out of processors quickly and cleanly. That is why Silvester is looking at new transmission algorithms for optical fibre.
"We can see the end point [for bandwidth on physical fibre]," he says. "The only way to increase fibre's carrying capacity is to come up with new algorithms."
Similarly, with the move to mobile working, the airwaves are already jammed to capacity in many cities. So Fujitsu is exploring ways to multiplex Wi-Fi signals so that a channel can carry more data.
But the most important trend is the coming shift to transaction-based computing, as exemplified above.
Great idea, but what's in it for Fujitsu?
Silvester believes the world is becoming more collaborative and multi-sourced. As Apple's App Store shows, there will always be a role for someone to collect, co-ordinate, broker and control the quality of data, apps and their processing. As the world relearns the disciplines of the mainframe, that is Fujitsu's future, he says.