Custom Applications using JMP and JSL: Why Predictum Inc.

Our programmers, engineers and statisticians have deep experience in the demanding world of research and manufacturing. We apply our cutting edge knowledge to constructing applications that allow your analysts and operators to take advantage of the most advanced analytical methods.

We develop in short cycles – often as short as weekly or bi-weekly, depending on your suitability, delivering value at the end of each cycle. This means that you start getting benefits from the project very early on in the process.

Typically, we begin, as most projects do, with requirements planning that details your objectives. We then create a development strategy and, with your approval, begin.

One of the biggest impediments to developing new systems is client participation. While clients recognize the need to have these systems built for them, they fear the amount of time and effort that they will have to put into their construction. Our development practices generally require only ½ to 2 hours per week of client involvement.

We know what you are up against as we’ve all spent a number of years working in production and research. Furthermore, our broad client base, in semi-conductor, life sciences, automotive and chemical industries give us familiarity with a wide range of situations. Our depth of experience allows us to be able to fully scope out the project ahead of time. We know what questions to ask and what details to acquire.

Rather than sketch things out on paper, where the dynamics involved cannot be fully understood, we can quickly develop key components like graphical user interfaces (GUI) and reports and put them in front of you for comment and approval. Further, we fully exploit capable development technologies that ensure our productivity.

We know that some companies are extremely concerned about exposing their data outside their organization. If this is your concern, we can develop without retaining any of your data for development and testing purposes. All we need to know is the structure of your data. Site tests, where we install the application on your systems, will verify functionality with your data.

To ensure progress and focus, we set up the infrastructure at the start of each project: (version-control (Subversion), bug-tracking (Trac) and forums). This combination allows you to easily access to the latest version of the project. If appropriate, we’ll integrate bug-tracking into the project, making it possible to gather and respond to bugs and feedback.
We also make ourselves available for any questions that might come up by way of email, Skype or video conference.

Often as much as 60% of a project will consist of functionality that we have already developed and deployed elsewhere. (None of this functionality is proprietary). Our library of these functions is growing all the time. The ability to incorporate pre-fabricated, field-tested code will shorten developmental timeframes and assure accuracy.

One of the reasons we can pull functionality from elsewhere is that we develop to accommodate a range of situations. We call this ‘abstract development’, which is the opposite of hard-coding. Developing in this way means we can construct a broad range of situations to develop and test. We can also test against potential situations not represented in your current data.

When functional changes are required, the current system is not tangled in a web of hard code. Often substantial functional changes can be added with little effort and little risk of introducing bugs along the way.

Because we are developing early on to accommodate foreseeable situations when they arise, no changes are required.