The SDK is part and parcel of AtScale’s move to a more agile quarterly release cycle. The platform also improves analytical query and security capabilities.
Today, AtScale is announcing their AtScale 2019.1. platform update, the first of what will now be four quarterly releases in a more agile development cycle. This release focuses on expanding support for enterprise data platforms like Microsoft Power BI, and includes, for the first time, a software development kit [SDK] that partners can use.
AtScale, a well-funded late-stage startup that was established in 2013, sees itself as occupying a unique position in the Big Data and cloud transformation space.
“Back around the 2010 time frame, I was running data for Yahoo,” said Dave Mariani, one of AtScale’s cofounders and their Chief Strategy Officer. “We were doing Big Data before it became part of the language, serving a lot of ads and web pages. The data got to be too big for the standard relational format. The tool of the day was Oracle RAC, and it couldn’t capture all the data we were collecting and make it work for the business. I also had users who loved different tools like Tableau, Qlik, MicroStrategy – or Excel, where all the data ended up anyway. Hadoop is a data lake and it allowed us to capture and store all the data. But Hadoop didn’t talk to these BI tools, and we struggled loading copies of the data into all these individual tools. Everyone also had their own way of calculating what a page view or an ad impression was, so nothing was consistent. And we had 200 engineers making big data small, sending it all over the world – where it was insecure.”
Solving these problems led to the founding of AtScale.
“Our goals were that we didn’t want to move data, that we wanted to speak to all tools, and that we wanted to use the same language,” Mariani said. “So we created this – a virtual data warehouse. Our virtual semantic layer has one view of the data, and lets each business have its own visual interface.” Mariani said that this means that they can work with the BI vendors rather than compete against them.
“We consider our main competitor to be the status quo – organizations who want to keep data small to make it work for their own system analytics,” Mariani indicated. “So Tableau users, for example, who create analytics that only work with those dashboards, We are really a general platform that does things like fraud detection in financial services, or in health care, doing fraud analysis at the claim level and measuring the effectiveness of procedures across their network.”
Their channel consists of global systems integrator partners like Deloitte, Accenture and Cognizant, and multiple sets of vendors – business intelligence companies, public cloud vendors and data platform companies like Snowflake, although the latter two groups can overlap.
“With us, the BI partners have access to more data in real time,” Mariani said. “We make Google Cloud’s BigQuery more effective, and we let them use Excel native, which no one else can do. We allow the data platform players to migrate from on-prem to the cloud. The systems integrators use us to help customers make these transformations.”
This AtScale 2019.1. release focuses on increasing the capabilities of these strategic partners to more effectively address hybrid data management and non-disruptive migration and portability to the cloud.
“We have worked with Microsoft to develop a native MDX connector for PowerBI, which is becoming more popular in the enterprise because it’s free, and because Microsoft is promoting it heavily,” Mariani stated. “We already had the ability in PowerBI to talk to us in SQL in tabular mode. “The addition of full MDX and multidimensional support provides the ability for PowerBI to leverage our Universal Semantic Layer, and see it in the same way, with everyone speaking the same language.” This frees business users from having to continuously re-define business logic in Power BI. It also enables live connectivity for data stored in Snowflake and Google BigQuery without the need to move data, through new, secure, governed data access for Microsoft Excel.
“We consider this to be a major feature enhancement,” Mariani stressed.
“The addition of support for SAML 2.0 on top of our patented TrueDelegation capabilities is another big one,” he added. “It expands our enterprise identity support to let users interact seamlessly with this, as well as with Active Directory and LDAP.”
The platform’s analytical query capabilities have also been enhanced.
“A key one here is the improvement of support for decimal data types, to allow for more precise queries,” Mariani said. “That’s especially important for retailers, where the numbers are now so large that they need to be exact.” Other improved query and management capabilities include enhanced time-series analysis with wide table planning, updates to AtScale’s Design Center for cube users, and new canvas management panel for version management.
The availability to partners, for the first time, of an SDK for AtScale’s platform is another big advance, Mariani stressed.
“You have to connect everything to you, and you have to connect to everything, and this will drive new ports to new back ends,” he said. “Before, we did all this internally, and we would add support for a new platform as part of a core release.” This release, for example, adds platform support for Teradata and PostgreSQL.
Going forward however, Mariani emphasized that the need for a more agile development cycle means that this way of doing things had to change.
“With our move to a quarterly release cycle, we wanted to make sure we could decouple data platform support from the quarterly releases, so we could be more agile,” he said. “This lets partners do the platform support themselves.”
AtScale 2019.1. is available now.