Knime updates analytics platform with new GenAI capabilities
The vendor's latest release improves the accuracy of its GenAI chatbot, including citations so users can know the source of the chatbot's responses, and features a new UI.
New generative AI features highlight the latest Knime analytics platform update, including improved responses from the vendor's chatbot.
In addition, Knime Analytics Platform 5.2 -- launched on Dec. 7 and now generally available -- features support for Azure OpenAI Service from Microsoft and a new UI aimed at making it easier to navigate the vendor's set of tools.
In fact, though not as dramatic as new generative AI capabilities, the simplified UI is just as significant a development by the vendor, according to Mike Leone, an analyst at TechTarget's Enterprise Strategy Group.
"To be a platform that can enable all folks to analyze, experiment [with] and explore data, the UI is important," he said. "Improving it was necessary."
Based in Zurich, Switzerland, Knime is an open source analytics vendor whose platform also provides data science capabilities.
Version 5.1 of the vendor's platform was released in June. Like version 5.2, it featured both generative AI capabilities and improvements to its user interface.
Shortly before that, in April, Knime launched an extension of its Business Hub named Continuous Deployment for Data Science that makes it easier and more secure for organizations to deploy and update data science projects.
New capabilities
Knime's new UI was developed in response to feedback from users, according to the vendor.
The vendor's analytics platform features thousands of nodes, which are individual tasks that can be executed using the vendor's platform. Among other things, the new UI aims to make it easier for both new and experienced customers to find the nodes they need to do their work, surface documentation related to the node and request help.
Given the simplified UI, more people within organizations might be able to use Knime, according to Leone.
Studies have shown that analytics use within organizations has been stuck around a quarter of all employees for well over a decade, in large part due to the complexity of analytics platforms and the training required to learn them. Any steps vendors can take to simplify the use of analytics tools are therefore significant.
"Now, businesses can scale Knime with confidence knowing it better empowers new and/or non-technical folks to explore the capabilities of the platform and do more with their data," Leone said.
Similarly, the new generative AI capabilities have the potential to boost analytics use.
Knime first introduced the preview of an AI assistant named K-AI -- the result of an integration with ChatGPT -- in version 5.1 of its analytics platform.
Like generative AI chatbots such as Copilot for Fabric from Microsoft and Tableau GPT from Tableau, K-AI enables customers to ask questions of their data in conversational language rather than code and receive responses in conversational language. Users can then ask follow-up questions, also without having to write code, to dig deeper into their data and hone the responses.
At the time Knime unveiled the chatbot in June, the vendor cautioned that K-AI was a prototype and not yet ready for production.
Nevertheless, in its early stage, K-AI showed Knime users how generative AI chatbots can broaden analytics use within organizations by greatly reducing the need to write code.
Businesses can scale Knime with confidence knowing it better empowers new and/or non-technical folks to explore the capabilities of the platform and do more with their data.
Mike LeoneAnalyst, Enterprise Strategy Group
For that same reason, many other analytics vendors -- for example, ThoughtSpot with the Sage tool it unveiled in May -- have made generative AI a focal point of their product development throughout 2023.
In version 5.2 of its analytics platform, Knime released an updated version of K-AI aimed at improving the accuracy of the chatbot's responses. In addition, the tool now cites the sources it uses to generate responses, which increases the transparency of query responses and enables users to easily check their accuracy.
"The ability to not only welcome new users but broaden the way all users interact with the platform extends to the new capabilities of K-AI," Leone said, noting that the new UI and generative AI both have similar goals.
In addition, the transparency provided by citing sources used to generate responses is significant, he continued.
"There has been a major push in the industry over the last several months on the importance of transparency and accountability. So I love the initial steps Knime is taking with capabilities to improve the accuracy and reliability of a response, specifically the ability to cite sources," Leone said.
Beyond enabling natural language processing, K-AI also now includes text-to-code translation capabilities and can suggest Python scripts and Apache ECharts visualizations to assist developers.
Meanwhile, in addition to the new UI and improvements to K-AI, Knime's latest analytics platform update features the following:
Support for Azure OpenAI, a tool that enables developers to access Microsoft's generative AI tools so they integrate those capabilities with their own to build generative AI applications.
A new UI for Knime's Python scripting node and a new Generic ECharts View node to improve the scripting experience for customers.
The road ahead
With a simplified UI and improved generative AI capabilities now part of the Knime analytics platform, the vendor would be wise to continue expanding the applications of K-AI, according to Leone.
It's important that the tool be more than merely a way to query data.
AWS recently unveiled Amazon Q, a chatbot that not only enables data exploration in QuickSight, the tech giant's analytics platform, but also enables conversational interactions in its contact center and supply chain management tools.
"I think there is a lot of value in continuing to invest and expand in how K-AI is integrated throughout the platform knowing that new users will naturally need an easy way to ask for help or guidance," Leone said.
Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.
