.svg "logo-sketch-development"){kind=logo}
A world leader in the field of in vitro diagnostics engaged Sketch to build an automated installer for updating the operating system and software of bacteria testing devices used in clinical applications.
The Company came to Sketch with
a problem: the current process for
updating their clinical testing devices
is manual, entails many steps (mostly
undocumented), and is inherently
error prone. Exacerbating the situation;
the only person that knows the manual
installation process has taken another
position. Even if someone else were
to be trained in the manual steps, the
process still represents a single point
of failure. Making matters worse, when
something goes wrong [often] during
the manual installation process, the
entire process must be started again
from the beginning.
If that wasn’t bad enough, the existing
process for building the devices’ base
image takes two full days to complete; it
installs six different software packages,
includes at least 10 reboots, and makes
Windows configuration updates.
What’s more, the manual process is
done quarterly to accommodate major
builds, plus another twenty more times
a year to create hot fixes. The Company
needs a new way to address their needs.
Knowing the key implementor’s departure was
imminent, the Company attempted to automate
certain steps in the process. This exercise led the Company to believe that automating the installation process end-to-end might not be feasible or even possible. At Sketch, we took this as “Challenge Accepted!”
The engagement kicked off with several days spent shadowing the individual implementing the manual steps. This allowed for us to build full documentation of the process. Once it was understood how the installation worked, Sketch developers were able to pinpoint steps that could be eliminated as well as to create discrete automations that significantly reduced the number of required system reboots.
To deliver the automation solution to the Company in a timely manner, Sketch leveraged their creativity to move the development of this automation into a virtual environment. Doing so shortened the feedback cycles for the developers and allowed frequent demonstrations of progress to the client. The delivered solution now includes two main benefits. Firstly, it provides clear status tracking and logging. Secondly, it includes an alerting system that, should an error arise in the automation, it pinpoints the section of code where the required fix will need to be addressed.
At the end of the ninety-day engagement, a manual, error-prone, 100-step, multi-day process was reduced to a single button push that no longer required an individual to oversee (or run) a manual process. Working in an iterative fashion with tight feedback loops the Sketch team was able to show progress – even at two weeks into the engagement.
The client shared that Sketch’s collaborative approach, fresh insights and curiosity helped them to see other ways of getting things done and felt their group improved as a direct result of our working together.
We hope these are the kinds of outcomes you were looking for. If you want to learn more about the work that went into this project, we're happy to answer your questions.
Better yet, let's talk about whichever project you were mulling over that brought you here. We love to talk shop, and the first few conversations are always free.
