The Quality Swiss Army Knife

  by Charles Oldham | 14 Aug 2007 | Permalink | Home Page, IT Management, Software, Software Business Models, User Experience

Originally I was going to comment on Allan’s excellent summary of our new provenance features, but it turned into a post in its own right. Whilst I think that the features provided by provenance will clearly be of use during pattern development I think they have wider scope than Allan feels they will.

The mindset I have had during development of (for the want of a better phrase) the “Saturn Ethos” is similar to what I would expect in scientific research; whilst the end result is important of equal importance is being able to prove that you can trust the result. In a piece of science research this is achieved by having your methods open for peer review and qualifying the answers with appropriate error estimates.

In our new release we strive to achieve this by being able to drill into the heritage of the data. The data we are dealing with is considerably more complex than scientific measurements so approaches such as error estimates have no direct parrallel, but I hope enough information is provided to allow end users to make a judgement as to the end result accuracy. At the end of the day you should reasonably expect to understand where each item of data in our model comes from.

Why I think this has wider scope is that frequently our end users will be taking executive action based on our end results, directly or as we feed knowledge into a wider ecosystem. A project sponsor using Tideway Foundation derived data needs to know that they can trust the data to take that action; that it is a real result that they will not be challenged on.

I have seen in the past as projects supported by Tideway Foundation establish themselves in our customers and reach a wider audience each new population of stakeholders ask the same questions about the data; especially due to the way our data can sometimes cross knowledge silos. It is natural to be suspicious of new data sources when they are in conflict with your existing ones you understand.

I believe there are additional benefits in supporting end user organisations gain trust in using Tideway Foundation in thier network. The provenance features also allow an understanding of why various machines were accessed and when that access occured. Black box systems are simple, but frequently feel uncontrollable to the end user. By providing greater transparencyof this our end users feel more comfortable that Tideway Foundation is not a black box but instead a system that is open and controllable.

There are additional benefits to operators of Tideway Foundation that fall out of our Directly Discovered Data structures, but those should be the subject of a future blog.

I believe we have not built a single complex tool to help pattern development, but a Swiss Army knife feature that will benefit ourselves, partners and end users of our new release of Tideway Foundation. I suspect at different phases of familiarity with Tideway Foundation it will be used more or less frequently, but quality always matters. It is sustained quality processes that underpin the reputation of major brands and I hope we will achieve the same.