Dundas BI

Dundas BI is a solid platform: it's more than a tool. With solid row-level security, a hefty selection of visualizations and the opportunity to present our user base with the ability to play with data, they won our vote. We compared them against: Tableau, Sisense, Domo and iDashboards. These comparative products had strengths as well, but when we made our decision, it was because Dundas gives our end users great control over visualizing what they see without unnecessary programming as well as our ability to use native SQL views we've already written with no need for modification. This in tandem with row-level security that we control at the data level made a perfect fit for our organization.

I like to get up and going intuitively. There wasn't a quick-start guide on how to create your first dashboard: it was pieces of learning. Having now gone through the majority of their 100-level and 200-level courses and having created a number of assets, I can see why Dundas wants you to do their training...but for heavy programming go-getters, I wish there had been a quick document to start. They did, however, respond amazingly as they hit my demands for learning head on and answered my questions rapidly, even assigning me access to a technician for a few days while I was in crunch mode.

The drag and drop feature for creating views is a very productive feature. The ETL is also great, it's easy to use and, for more advanced users, it evens permits to write direct sql against tha data source you choose. Regarding performance it has two possibilities, you can cache in memory or in DBI data warehouse. The collaborative feature of this tool is a must to have feature, I specially like the data driven notification option that could alert you by email when a specific metric goes behind a threshold or state.

It needs to take a step further and aboard predicted analytics!

Dundas seems to really understand the need of handling big data and the tools needed to process these overwhelming mountains of 1's and 0's into actionable data. While working with the demo of Dundas I was able to create a very helpful tree-map that showed me the dissection of purchases comparatively from vendor to vendor to evaluate our performance or "loyalty" per vendor by item category and so on.

The U.I. of Dundas, while quite robust, seems to concern me in regards to the amount of boxes, checks, and clicks and long clicks I have to make to merely navigate my way through the program. I would rather see more keyboard shortcuts and less "Windows Vista" popups that slow my work. I felt like the settings I would implement would get lost. For instance, I would filter the vendors that weren't pertinent to my particular evaluation, then also the item categories that I wasn't interested in working with. Later on, when in the final building of the dashboard I wanted to add a checkbox filter to allow the end user to select individual item categories and be able to view 1 or all 20-30 at a time. Strangely, even though I filtered (or so I though) this data in the data cube level, I realized that the checkbox list populated with ALL the categories. This was quite frustrating. Also, I have yet to figure out how to go from an unfinished dashboard back to the data metrics screen where I built the KPI to start with. Unless i'm mistaken, it seems you have to go back to your saved metric set and then create a new dashboard after edits. Hopefully I'm wrong in my assessment of this but even so, the answer did not seem to be intuitive .