How to maintain data quality in Insight

Question

Hi

We are a small in-house IT staff team. We are thinking about using Insight for better control of our IT hardware like computers and such. I recently tried Insight in Jira, but there is a potential pitfall I am worried about.

If everyone in the team can register new assets (objects) in Insight, there is a chance that someone eventually will enter something wrong in the attributes, like a typo for example. Since a lot of the attributes are text fields.

We would like to avoid using spreadsheets because someone will eventually enter something wrong, and the whole data quality is unreliable.

Since I am not that familiar with Jira, can someone please advise how to go about this?

Usually, in a CMDB you would be able to add products beforehand. After that, you can assign CI's or add objects to the CMDB. But since Atlassian Insight boasts about not being a classic CMDB I am curious about how they intend us to do this.

How do we maintain data quality in Insight and avoid typos and errors in text fields?

Mikael Sandberg · Answer

There is nothing in Insight that prevent you from adding typos, same thing as in Jira there is not feature that would prevent you from doing the same. Hence my suggestion that you could use a request form to create new objects and have a review process before the object is actually added, either manually or via automation.

joobjoob · Answer

Hi

Sorry but I think you misunderstood the question here. I already know about the import function, but thanks for mentioning it. However, that was not my concern.

Is there any way to mitigate typos and errors when employees add objects to Insight? As you can imagine a lot of text attributes leave room for mistakes and degrades data quality.

Is there any way to avoid this somehow?

Mikael Sandberg · Answer

Hi @joobjoob,

Welcome to Atlassian Community!

You can definitely use Insight as a CMDB, but you can use if for so much more and that is why it is boosted as not being a "classic" CMDB. Currently the importers are limited to csv, json, and Insight Discovery which is an agent-less discovery tool. You could also use external iPaaS tools to import objects, we are using tray.io to import our assets in Jamf and InTune into Insight and we are looking at doing the same using Kandji instead.

You can also use automation to create new objects and have a hidden request form that your team uses. You could then have a review process that happens before the asset is created in Insight in order to avoid typos.