The world has seen an explosion in data with the incomprehensible number of 2.5 quintillion bytes of data being produced every single day. Much of this data is semi-structured or unstructured data, stemming from the content produced on social media platforms in the form of pictures, videos and text.
Many organizations use JSON files to store semi-structured data effectively. When businesses want to analyze this data together with their structured data and form an integrated, 360° view of their customers, products, suppliers, and so on, they need to bring JSON files into a table structure.
What’s the challenge?
This process can be complex and time consuming. It typically requires data preparation in the form of ETL (Extract Transform Load) workflows that transform the data from its current JSON format into the table format that allows users to join data to existing tables and schemas in the database and query it using BI and analytics tools as well as SQL queries.
Dealing with JSON files in workflows or third-party tools when the rest of your data is already in a centralized database also introduces inefficiencies and the potential for additional steps each time changes need to be made or new sources are introduced.
Exasol makes JSON functions native
We want to empower our customers, partners and users from our community with the fastest and most seamless experience when they work with data to address their business challenges and find answers to their questions.
To address the need of organizations working with semi-structured data, we have made JSON functions executable directly in the database. When you work with JSON files, you no longer need to use user defined functions (Loading...UDFs) to execute JSON functions. Instead, they are available directly through SQL.
The time saved by removing additional steps from the data preparation process can open up capacity for you and your team to address other key topics for your organization’s data strategy. These may range from data security to effective data democratization and potential implementation projects for new solutions.
By giving you a standardized, simplified and highly effective way of dealing with semi-structured data directly in the database, we want to give you some time back and also help you deal with all your data in one place.
What this means for you
Working with data should be fun. Data analysts and data scientists spend far too much time preparing, cleaning and manipulating data to get it into the right structure for their analyses. But most of them would much rather spend that time on analysis, research and investigation to find insights that are truly valuable for the business and improve the organization’s products and services.
Adding data sources into your existing model should not be complicated or time consuming, so we want to make this experience an easy step that doesn’t interrupt your flow.
When you work with JSON files in Exasol, you can go right ahead and work with them directly in the database.
As a result, you will enjoy faster performance of your queries when processing JSON files. You can also save time during the data prep process because there is no need for complex ETL workflows or additional up-front modelling to handle your JSON data. And the native support for JSON files gives you more flexibility to introduce data sources you weren’t able to integrate previously or to add new data sources you’ve been wanting to analyze.
Where you can take this
Remove the headache of dealing with semi-structured data and gain more insights from the data sources you have invested in. With Exasol you can go even further. Once you have your JSON data in the database, why not run sentiment analysis on your customers’ reviews of the products in your online store or the services you provide? You could even train a Tensorflow Model directly in Exasol with GPUs.
Eva Murray, Technology Evangelist, Exasol