PowerBI has become something of an industry standard when it comes to building data dashboards of the type beloved by CIOs. But what if your data is already in AWS? QuickSight is a lesser-known analytics tool. It integrates with a multitude of different AWS-based databases and data stores. And the dashboards are friendly, easy to set up, and lend themselves to really flexible reporting.
There's also a heavy dose of Generative AI making its way into QuickSight. AWS like to call this "Generative BI" (as in, generative business intelligence tools). What this actually means is that you can type out some simple instructions to QuickSight and have it build you a chart that illustrates the trends you want to see. Further down the line, you'll be able to ask it to build a whole presentation (in the
form of "stories" that feature words interspersed with charts). Imagine the time saved! The demos are incredibly impressive, although as with all things GenAI, there'll inevitably be some fine-tuning and practice required to get perfect results.
Despite all this, it's still pretty hard for small and mid-sized businesses to make their data really shine. For a start, most companies have loads of different systems in play, and siloed data could be spread across any number of these. Quite often the systems won't talk to each other - or worse, you might end up with the same data in multiple different tools, where it can become stale, inconsistent or out of sync.
We've thought about this a lot. What we think businesses want is a way to make data-driven decisions, based on all the relevant data, without having to build hugely complex data warehouses, which then need constant TLC to keep pace with all the data in all those tools. They may not have the budget to employ a team of data scientists, and they certainly don't want to have to engage those people whenever a report needs tweaking.
Much better to have a dashboard that non-technical users can play with -- adjusting the charts with an easy interface, while still harnessing powerful data models underneath.
That's surprisingly tricky to achieve, and QuickSight hasn't fully nailed it yet. What we've come up with is a combination of AWS' tooling with our own data science experience. We deliver this as a package, using runbacks we've developed. But that's mixed with workshops and consultancy so that we can understand what you want to do with your data. So you'll end up with a tool that gives you the insights you need, but is flexible enough to adapt as your business grows.
The data integration layer uses Amazon AppFlow. This hooks into several well-known SaaS products and databases. For customers that use unsupported SaaS products, our development team can build custom integration modules; and because we've based our platform on AppFlow, we're able to save time and deliver the integration quicker than if you were building something from scratch.
We then use Amazon S3 to store the underlying data. This means the storage grows as your data does - there's no need to worry about running out of disk space. Amazon Glue transforms this into a space-efficient columnar data format (known as Parquet), and Athena allows us to query the data using SQL.
When building the QuickSight dashboards, we'll join the data across your different SaaS tools, using SQL joins. We will provide you with a set of well thought-out dashboards, but you'll have the ability to adjust these and create new ones as your requirements change. We provide an ongoing managed service to help you with the data plumbing underneath.
I really like this combination of great tooling and good engineers. It means we can make best use of our talented team, but free them up from mundane, repeatable tasks which are best left to a managed service. And it means you get a flexible product at a lower price, delivered quicker than completely from-scratch internal tooling. Plus, it's tried-and-tested and comes with our decades of combined consultancy experience, too.
