The cloud is full of databases. More specifically, the world of cloud computing is populated by a massive number of cloud databases, many of which belong to the increasingly popularized notion of the cloud data warehouse.
A cloud data warehouse (and indeed, a data warehouse back down on Earth located on a mainframe or some other pre-web notion of mass-scale computing) is a means of creating a wider, larger and more expansive repository for information than can be built with one single database or source/stream of data.
Used to coalesce a variety of an enterprise’s operational information systems, a cloud data warehouse can support both analytical and transactional data processing disciplines and tasks as it works to serve various Business Intelligence (BI) functions, while it also handles data storage and a smorgasbord of other data management practices.
Built to capitalize upon the breadth and power of cloud services, a cloud data warehouse differs from an ‘old fashioned’ data warehouse by virtue of its capability to offer elasticity, service flexibility and public cloud scale.
Firebolt: hot enough to melt a cloud Snowflake?
So cloud data warehouses are here and this is a burgeoning marketplace. Established in 2012, Snowflake made big financial news in this marketplace last year with its IPO. Now securing US$37 million in series A financing as reported here on Forbes by Gil Press, a potential Snowflake melter has also emerged in the shape of Israel-headquartered Firebolt.
Co-founders at Firebolt Eldad Farkash and Saar Bitner suggest that the global cloud data analytics market is expected to grow to some US$65 billion by 2025. Yet, they contend, current cloud data warehouse deployments struggle with performance and high costs, limiting their ability to address the needs of enterprise operations and customer-facing analytics. They explain how they ‘realized’ that with terabytes of data, existing cloud data warehouses could not scale to deliver the performance and efficiency organizations need to power modern Business Intelligence (BI) tools and analytics.
Farkash and Bitner claim that Snowflake and some of the other bigger name cloud data warehousing platforms (Amazon Redshift, Microsoft Azure and Google BigQuery) come from an era when the warehouse’s central USPs included file management and the initial challenges of building upwards to web-scale cloud-scale size and scope.
As a means of differentiating itself in the market, Firebolt has been established for a more ‘data usage’ analytics-centric world i.e. this is not about building a big enough data warehouse (it’s cloud-based, after all) or even knowing your way around it on a virtual forklift truck — this is all about making sure the data held in the warehouse’s wooden crates is more ordered and prepared for consumption in live business environments.
“While companies can store massive amounts of data, most organizations are only able to analyze a fraction of that big data and often find themselves looking at stale data that does not reflect the current state of their business,” said CEO Farkash. “For companies to flourish today, they need to move fast and they should not be forced to make data compromises to achieve only a small part of the business value that their data holds. With Firebolt, organizations can finally gain the insights they need without breaking the bank.”
Drinking from the firehouse: information suffocation
As we have explained before here, data ingestion is a massive challenge for organizations working with new cloud data platform technologies. Simply adopting cloud, pressing the ON button and drinking from the firehouse usually leads to drowning… or some form of information suffocation at best.
This ingestion and data preparation phase forms a key part of the company’s differentiating go-to-market technology proposition. Firebolt ensures that the data ingested from its customers is subject to sorting, data compression and other indexing processes designed to bring order to the chaos of modern transactional/operational IT systems.
It’s not that this tier of data management is new per se, it is more a question of Firebolt creating a cloud data warehouse service that applies the software intelligence needed for this process to happen automatically (and hidden from view) as a part of its live production environment tasks. This, along with a more heavily consumption-based (as opposed to subscription-based) pricing model is what sets the company apart claims Farkash.
What happens next in the cloud data warehouse space may be governed by the sheer mass of information that enterprises attempt to put in them. Although Firebolt is pushing for greater efficiency and order through automated data indexing, if we start to use data warehouses as the ‘dumping ground’ for more unstructured data (the sort we would normally expect to finding swimming in the waters of a so-called data lake), then the reality of the cloud data lakehouse may come to pass.
If Firebolt aims to be the super-efficient forklift for the cloud data warehouse, then it had better be working on its speedboat version for the future.