A paradigm shift is coming in data analytics. While traditional definitions of dimensional modeling, ETL, and data warehousing have served us well, new advancements are making analytics against non-aggregated data not only a reality, but full of practical benefits.
Transcript:
Welcome to zero gravity. And thanks for checking out Incorta. You know, we're in the middle of a transformation in both data management and data analytics. And I'm not talking about concepts like data fabric or data mesh. No, this is something more fundamental. I'm Cameron O'Rourke, part of the product team at Incorta. A couple of years back, I was out looking for a new car and I test drove a Tesla. It wasn't just the acceleration or the autopilot that got me. This was a fundamentally new approach. And right in that moment, I knew there was no way I could go back to driving a regular internal combustion engine, the old way, just no longer made sense.
So park that idea for just a minute and imagine if you didn't have to build a traditional data pipeline to serve your Analytics users? What if you could deliver blazing fast analytical query performance with transaction level detail without ever having to reshape or aggregate the data to achieve that speed? If we could eliminate the need to create ETL code, star schemas or other dimensional models, just like my Tesla, this would be revolutionary. And the good news is that not only is all of this available today, but we believe that this is how we will all be doing analytics in the future. So by way of example, Incorta is a cloud based data and analytics platform that offers a complete end to end system for acquiring, processing, analyzing and presenting operational applications data. Incorta takes in data from large operational applications like SAP and Oracle, and can immediately perform very fast analytical queries directly against the raw original data.
There's no need for ETL stages like bronze, silver, gold, just to get acceptable query performance. Instead, the data is left in third normal form with no data loss, retaining all of the original transaction detail and data fidelity. Now, this is extremely hard, especially for smart technical people to wrap their heads around. I mean, we've been creating dimensional models in star schemas for what like 30 years now. So how's it possible that a funky little startup delivered something totally new? Well, we've combined some familiar data lake and database technology with our own secret sauce. It starts with data connectors, and parallel loading into a data lake based on parquet, which is a compressed Columnar storage format. Then, as the data is being loaded, we pre process the data and create an extremely efficient in memory optimized join index. The data is loaded into memory in its compressed columnar form, such that we can handle 20 times the normal amount of data and read it extremely fast.
The query engine then leverages the data map, and is very smart about how it navigates even very complex queries. The results that our customers get are stunning. Queries that used to take hours are now returning in seconds. We hear this over and over again. And it's one of the main reasons that we have virtually no churn and that our customers grow their usage of Incorta every year. Now, the benefits of no ETL data prep of enabling analytics on bronze or original data, don't just stop at query performance. Analytics on detailed data means complete flexibility and traversing the data. And the ability to join data together even from different application domains, it now becomes possible to implement fine grained security that mimics the original application system, so that more data can be safely analyzed by more people.
Drilling down from high level metrics to detailed transactions is now trivial to enable validation, auditing, correction of source data and direct action on your findings. Finally, it means one copy of the data and a single source of truth for all workloads, whether analytics, operational reporting, or even machine learning. This paradigm shift disrupting the fundamental assumptions of the data pipeline itself is one of the biggest things to happen in data analytics in decades. So remember, my Tesla buying experience. Now, just like Tesla, Incorta doesn't have the size and maybe maturity of some of the big players in the industry but once you've experienced Incorta, fair warning, it's going to be hard to go back to your old technology. But don't take my word for it. I really invite you to sign up for a trial, connect to one of your application systems and just give it a go or even better sign up for one of our upcoming virtual hands on labs happening the week of June 1st. Enjoy the rest of zero gravity and thanks for listening.
.svg){kind=small}