Data is fluid. Information is fluid because it moves in between applications, services, functions, tools and the database systems it resides in. Information is fluid because it is increasingly channelled in real-time in contemporary data streaming environments - and information is fluid because it is constantly changing and morphing its form, value and structure across the many types of data that we now seek to manage in business.
Information is also fluid because we channel some of it into data lakes, a term used to describe the pool of largely unstructured data that a business may not make immediate use of… but does still identify as potentially valuable.
From the data lake we then get the data lakehouse, a hybrid term used to denote some of the structures we would find in a more ordered data warehouse with the expansiveness and lower cost functionality of the data lake. But, finding our way around the data lakehouse, even with its more defined edges and channels can still be tough.
Today’s businesses are collecting data from more sources than ever before, with volumes increasing an average of 63% every month in recent research. In an effort to consolidate, many companies have invested in data warehouses or data lakes and data lakehouse, using knowledge graphs to map relationships between data sets.
A codeless knowledge graph
Working at the fulcrum of all of these above technologies is Kobai, a company that describes itself as a codeless knowledge graph platform. The organization’s latest product is called Saturn, a knowledge graph designed to to harness the scale, performance and cost efficiency of the data lakehouse architectures. Perhaps somewhat strangely named (Saturn is god of agriculture rather than water, Poseidon may have been more logical), Kobai’s Saturn extends the capabilities of its core platform, integrating every use case and function into a single semantic layer.
“Our mission is to allow both business and technical users to quickly answer their own questions using their own terminology, without writing complex code,” said Parag Goradia, co-founder and CEO of Kobai. “By bringing lakehouse scale to the knowledge graph, Saturn empowers organizations to apply knowledge graphs as a foundational piece of their semantic data layer.”
Goradia and team say that business users need quick insights to make day-to-day decisions, which require connected data from across the enterprise. While knowledge graphs create crucial pathways between datasets, they are inherently difficult to scale. With Saturn, organizations can leverage the ease of knowledge graphs with the scalability of a data warehouse.
Integration, virtualization & collaboration
Powered by Databricks or Snowflake, new capabilities include direct integration. Embedded in the data layer, organizations can query data without moving it from the lake or warehouse, following W3C and Lakehouse open standards for complete interoperability.
There’s also on-demand and burstable compute using the underlying data layer for faster graph queries and ML training without virtualization. The company also boasts of good collaboration functionality i.e. users are able to publish business question as SQL views to integrate with existing data science and business intelligence tools
Knowledge graphs are generally built on more limited database systems or costly in-memory solutions, Kobai claims to be going beyond that point.
Built on the foundation of graph databases, Kobai’s codeless platform and new new Saturn knowledge graph work directly with Kobai’s Studio framework and its Tower visualization products to help data scientists and software application development engineers achieve the full value of their existing storage investments and meet the challenges in modern data.
Back-end burstable benefits
Fairly complex technology yes, but as with so many enterprise software platforms today we see a high degree of user automation and back-end control being tabled here. This talk of burstable compute boosts (just above) means software engineers will get more control and power to steer their boat [applications] across the business landscape [data lakehouse] hopefully avoiding rocks [cloud miscongurations] and port docking challenges [integration problems] to enjoy clean sailing.
The data lake and data lakehouse may now be more navigable, all we need now is a weather report.