Knime adds new UI, closer Snowflake link to analytics suite

A modernized user interface, a new environment for Python users, and deeper integration with Snowflake highlight the latest version of the Knime analytics platform.

Knime, an open-source analytics provider founded in 2004 and based in Zurich, Switzerland, launched version 4.6 of its platform in a blog post on June 15. A second update this year is expected in the fall.

The update unveiled last week is aimed at end users, including the many self-service users who have started using Knime’s analytics tools over the past two years, according to Phil Winters, strategic advisor at Knime. .

In the fall, the vendor is planning a release aimed at adding enterprise-wide functionality.

“This particular release has to do with analytical depth and breadth and usability for individuals,” Winters said. “There are a lot of things added for people at different levels. This release is primarily designed to help all different types of end users progress.”

For example, new visualizations can be consumed by any user, while the environment for Python users and integration with cloud data provider Snowflake is for more sophisticated customers.

Meanwhile, the vendor’s open-source approach underlies all versions of Knime, Winters continued.

“At the heart is this concept of openness,” he said. “Everything for the individual is open source, which means they can benefit from the community and benefit quickly because we don’t have to rewrite their code. The other thing, because we can never do absolutely everything ourselves, we have to make it as easy as possible to mix and match. »

Given the open source architecture of the Knime Analytics Platform, communication and connection between Knime and other systems is essential. To facilitate this communication and connection, the provider uses nodes, which are essentially connectors that allow its users to easily write, read and move their work.

It’s this network of nodes that makes Knime attractive to many organizations that operate multiple systems, according to Mike Leone, analyst at Enterprise Strategy Group (ESG).

“Knime doesn’t care how a client’s environment looks,” he said. “They recognize that customers have different needs and requirements based on their use of different tools in different environments. This level of flexibility is being embraced by organizations that have struggled to transform due to the sheer complexity of their massively distributed data environments.

An example screenshot from Knime shows the new Analytics Provider UI.

New Abilities

The first thing Knime users will soon see when using the provider’s analytics capabilities is a modernized user interface that is now in preview.

Knime’s current login screen — the launch point for all Knime workflows — is starting to look old, according to Winters. The new user interface will be web-based, and in addition to a new look, its node repository comes with stronger search and filtering capabilities.

“Our interaction screen for creating workflows was starting to look a little stale in the teeth,” Winters said. “He had fantastic abilities, but he was neither sparkling nor modern.”

Leone noted that the modernized user interface will particularly appeal to the growing number of Knime self-service analytics users.

The provider historically catered to data scientists, but recent updates have also made its platform more accessible to business users, and the new user interface continues that trend.

“While some of the traditional BI players are focused on expanding into the data science space, Knime is and has been there,” Leone said. “With Knime’s new modern user interface, they strive to provide a visual experience similar to traditional BI players, but with a much more robust data science offering.”

They recognize that customers have different needs and requirements based on their use of different tools in different environments. This level of flexibility is embraced by organizations that have struggled to transform due to the sheer complexity of their massively distributed data environments.

Mike LeoneAnalyst, Corporate Strategy Group

In addition to the modernized look, the new web-based user interface completes the decoupling of Knime’s front-end from its back-end and makes Knime a “headless” platform that allows users to compose their own analysis workflows.

Beyond the upgraded interface, users will see improved Python scripting capabilities and the ability to work with models directly within Snowflake.

Both of these features are now generally available.

Python is a popular programming language used by engineers and developers to build analytics capabilities. To better enable them to work with Python, the Knime Python extension – one of the nodes of the provider – now contains its own Python environment so that users can immediately start writing scripts without having to install new software first.

Additionally, when users create their own nodes using Python, they can now more easily share their creations with the rest of the Knime community by dragging and dropping them into the Knime Hub where users can upload and share their analytical creations.

The goal of the new Python environment is to add ease of use for end users and managers, simplifying coding and sharing while centralizing it in a controlled framework.

“Having pure Python Knime nodes will be incredibly valuable,” Leone said. “Users no longer need to leave Knime to use and integrate Python features. This gives technicians more agility by leveraging Knime for all orchestration, while sales reps gain the control they need. to improve Python code sharing and collaboration.”

Deeper integration with Snowflake also addresses ease of use.

Knime has integration with AI provider H20.ai that allows users to develop no-code/low-code machine learning models.

Prior to the release of Knime 4.6, vendor users could build predictive models using Snowflake data in concert with H20.ai’s low-code/no-code capabilities, but to do so they had to move their data out of from Snowflake to Knime, develop their models in Knime, then send their data back to Snowflake for storage.

Now, thanks to its improved integration with Snowflake, Knime users can develop and run H20.ai machine learning models directly through Snowflake without having to move their data back and forth.

The result is increased speed and flexibility.

More capabilities and future plans

In total, the latest Knime platform update includes over 10 new analytics features.

Beyond the new user interface and improvements to make it easier to use Python and Snowflake, the update includes new visualization nodes for building and exploring data apps, improvements to the Knime database that allows users to work with data where it resides by building SQL statements in Knime and easier interaction with Microsoft’s Azure Synapse Analytics.

Knime’s next update, scheduled for fall 2022, will focus more on the enterprise than the end user, according to Winters. No details are available yet, but Winters noted Knime’s hub will be featured prominently.

Knime’s hub has grown significantly over the past two years with around 10,000 user-created analytics assets now available, up from around 2,500 in 2020.

“When people use them, they want to be able to use them within their business,” Winters said. “They want to be able to have teams that share them and collaborate. So what you’re going to see coming from us in the fall are the first of our paid services so people can start watching all of this.”

Enterprise Strategy Group (ESG) is a division of TechTarget.

About Nunnally Maurice

Check Also

COVID can’t break South Africa’s love affair with malls

JOHANNESBURG, June 29 (Reuters) – Two days before it opens to the public, workers rush …