Snowflake Computing came out of stealth today and announced $26M in Series B funding. They hope to modernize the data warehouse by creating a cloud-based system that can process both structured and semi-structured data in a single system, giving their customers what they believe to be the best of both worlds.
The funding is led by Redpoint Ventures along with Sutter Hill Ventures and Wing Ventures. Sutter Hill funded the seed round and Series A. The company has been around since 2012, but is just emerging today announcing the funding and its first product.
The product, The Snowflake Elastic Warehouse, is a cloud-based service, what they are calling Data Warehouse as a Service. The data warehouse concept actually dates to the 1980s when companies such as financial services and credit card companies were starting to deal with lots of data coming from relational databases and they needed a way to pull data from various systems to make reasoned decisions. These early systems required tremendous resources and IT help to pull off, often took years to build, were expensive to build and maintain and complicated to run, requiring experts to help formulate queries and generate answers.
Snowflake saw an opportunity to modernize this concept by reducing the complexity and putting the whole thing in the cloud. That takes the hardware out of the equation because users can access as much compute and storage resources as they need at a reasonable price, and IT doesn’t haven’t to build or maintain a system in a data center.
End users don’t have to worry about anything except loading data and running queries and the system can scale up or down to meet their needs. This means they aren’t limited by the size of the data stores and they don’t need IT or DBAs to help them –but Snowflake didn’t stop there.
CEO Bob Muglia said that the old data warehouse relied on relational databases in neat rows and tables, but today’s data is a mix of structured data and semi-structured data. Instead of making customers choose one or the other or have tools to process them separately, Snowflake built its own database to process both types of data in a single system.
It’s a bold move in a time when companies are looking at open source solutions, choosing one way to process the data, and building a solution on top of that. Instead, Snowflake went its own way and built a system from scratch. Muglia says because they wanted a system that processed the two types of data, structured and semi-structured, it required building their own database to accommodate that vision.
The result was a single product that can understand both types of data and provide answers to queries in as little as 15 minutes instead of hours or days under the old-style data warehouse. Snowflake also stuck with the older concept of data warehouse, rather than latching to the large data set buzz word du jour, big data. That’s partly because they don’t care about the size of the data set. Users can provide as little or as much data as they wish and Snowflake will process it for them.
Photo Credit: (c) Can Stock Photo