5 New Analytics Features to Accelerate Insights and Automate Decision-Making

Adam Hughes

#MongoDB World

The applications we use every day are continually delivering richer experiences and working more efficiently. One of the driving forces of this progress is analytics. As organizations ingest and use ever increasing layers of data, they are able to derive more timely insights about their users’ preferences, patterns, and needs to deliver just-in-time information and choices within their applications.

The next generation of applications will take a huge leap in intelligence by integrating real-time analytics into their app experiences. Such analytics will increasingly be automated, developer-driven, and incorporated seamlessly within data platforms alongside transactional — or application workloads.

As announced at MongoDB World 2022, MongoDB will introduce five new features this year that will help businesses modernize their analytics: Column Store indexes, MongoDB Atlas Data Federation, MongoDB Atlas Data Lake, MongoDB Atlas SQL Interface, and distinct tiering for analytics nodes.

Using these features will automate decision-making and drastically decrease the time it takes to get application insights in front of users.

Modernizing analytics around operational data

Today, in order to create dynamic in-app experiences, businesses need to take multiple steps — collecting application data, sending it to a data warehouse or data lake to run analytics on it, deriving insights, coding new experiences, and releasing the app back to users.

Modern applications must be able to automate this process by capturing and processing the data at the source — that is, in the application. The data inside your application is the most valuable and current picture of what is happening with your business.

Combining real-time, operational, and embedded analytics — what some call translytics, HTAP, or augmented transaction databases — now enables analytics driven by application data to help determine, influence, and automate decision-making for the app and provide real-time insights for the user.

Real-time analytics is, as the name implies, done nearly instantly, usually on data that resides in an application. Examples include fraud detection for banks and personalized offers or recommendations on an e-commerce site. The analytics can range from basic aggregations to machine learning models that provide insight and automate an action, such as sending an offer. One example is Ticketek, an Australia-based event ticketing company, which uses real-time analytics to make critical decisions, such as whether to open up more sections of a venue or put on more shows.

Operational analytics is the process of finding insights from your data sources to improve decision-making for the daily operations of a business. Use cases include real-time reporting, improving overall operations, and product analytics. Online grocery Boxed, for example, was able to manage inventory levels during peak demand thanks to real-time data and insights directly from MongoDB Atlas.

Embedded analytics enhances applications by embedding data visualizations and dashboards with MongoDB Atlas Charts, providing users with relevant insights when and where they need them.

What's New

Here are five advances announced at MongoDB World that can help businesses modernize their analytics:

Column Store indexes: This feature enhances analytical queries by allowing developers to deliver real-time analytics on live, operational data. It also improves the performance of common analytical queries by adding a structure on top of collections that groups similar fields together to speed up reads. This eliminates the need to offload analytics to disparate specialized systems and rely on complex and fragile ETL pipelines that ultimately slow down the time to gain insights.

Atlas Data Federation: Atlas Data Lake is relaunching as Atlas Data Federation to reflect our focus on the value of federation. MongoDB Atlas users have the ability to query several data sources at once.

Atlas Data Lake: The new Atlas Data Lake provides a cost-effective data store optimized for high-performance analytics on large volumes of data. Atlas Data Lake delivers analytical workload isolation, allowing you to perform complex, long-running, or large analytical queries without impacting your production application. Fully integrated as part of the MongoDB Atlas, Atlas Data Lake can be provisioned alongside your Atlas Database, making the ingestion and optimization of data simple, with no infrastructure to set up or manage.

Atlas SQL Interface, Connectors, and Drivers: Atlas’s new SQL capabilities allow people who mainly work in SQL tools, such as data analysts, to easily interact with Atlas data. Users can query Atlas data via a BI tool or SQL driver and are able to directly query live data and gain enhanced schema control.

Distinct tiering for analytics nodes: Users can choose an appropriately sized node tier dedicated to their analytics workload without needing to change the tier of the entire cluster. This can enhance the performance of your analytics workloads; you can provision only what you need if your analytical workload requirements are less than your transactional requirements.