Getty Images

Qlik's Kyndi acquisition targets unstructured data, GenAI

The longtime analytics vendor's latest purchase adds support for unstructured data that can be combined with structured data to inform GenAI models and improve decision-making.

Qlik on Thursday completed the acquisition of Kyndi in a move that adds support for unstructured data so that users can more comprehensively train AI models and applications and inform other data products.

Financial terms of the transaction were not disclosed.

Qlik, based in King of Prussia, Pa., is a longtime analytics specialist that in recent years has used acquisitions to expand into data integration and AI -- including generative AI -- development.

Among its more significant acquisitions were Podium Data in 2018 and Attunity in 2019 to first add integration capabilities. Qlik acquired in 2020 and Talend in early 2023 to expand its integration platform, and Mozaic in late 2023 to add decentralized data management.

In addition, Qlik acquired CrunchBot in 2019 to add an AI chatbot and Big Squid in 2021 to improve its automated machine learning tools.

Meanwhile, Kyndi in San Mateo, Calif., provides natural language processing, search and generative AI capabilities as they relate to unstructured data.

Prior to its acquisition by Qlik, Kyndi had raised $47.4 million in funding.

Additive capabilities

Qlik Cloud, the vendor's combined data integration and analytics platform, was designed to enable customers to work with structured data such as financial records and point-of-sale transactions when developing data products.

It does not, however, support unstructured data, which includes text, audio files, video and more.

Qlik's acquisition of Kyndi adds that missing support for unstructured data, enabling customers to combine disparate data types to develop more comprehensive data sets that can be used to better inform data models and applications, including generative AI models.

Generative AI models, in particular, require vast amounts of data to deliver accurate outcomes.

Unlike traditional AI models, generative AI models respond to queries regardless of whether they have been trained on the data relevant to that query. Sometimes those responses are so far afield that it's obvious to a user that the output is incorrect. Other times, however, the response -- while incorrect -- is plausible enough that a user might be fooled and make a decision based on faulty information.

This will enable [Qlik] to extend more into the unstructured data space. This is something organizations are increasingly interested in -- they want more context from unstructured data ... because there's simply so much of it.
Mike LeoneAnalyst, Enterprise Strategy Group

Those AI hallucinations can have dire consequences if not caught, making it imperative that generative AI models be trained on as much quality data as possible to increase the probability of an accurate output.

Adding unstructured data to structured data improves the training of models and increases their likelihood of delivering correct results.

As a result of Qlik's previously missing support for unstructured data, the acquisition of Kyndi is significant for Qlik, according to Mike Leone, analyst at TechTarget's Enterprise Strategy Group.

Data platform vendors such as Databricks and Snowflake, which like Qlik have made generative AI a significant focus of their product development plans over the past year, already enable users to combine structured and unstructured data.

"This will enable [Qlik] to extend more into the unstructured data space," Leone said. "This is something organizations are increasingly interested in -- they want more context from unstructured data ... because there's simply so much of it. Customers will be empowered to integrate and analyze structured and unstructured data together, which will enable even better insights."

Leone noted that Qlik offered a strong suite of AI and machine learning capabilities -- including generative AI -- before its acquisition of Kyndi. In September 2023, the vendor introduced Staige, a suite that enables customers to access Qlik's traditional AI, generative AI and machine learning tools in a single environment.

Kyndi's acquisition, therefore, stands to enhance those preexisting capabilities by adding the insights that can be gleaned from unstructured data.

"Customers are already seeing benefits [from Staige] by implementing their own generative AI models to better inform decision-making with a trusted data foundation delivered by Qlik," Leone said.

James Fisher, Qlik's chief strategy officer, likewise noted the importance of adding support for unstructured data through the acquisition of Kyndi.

"While Qlik had already embarked on the journey of leveraging AI to drive insights on structured data, the addition of Kyndi's technology [accelerates] our momentum by adding unstructured data," he said. "Kyndi ... expedites our capacity to incorporate diverse types of unstructured data, propelling us beyond our current capabilities."

Once integrated, Kyndi's capabilities will allow Qlik users to easily blend structured and unstructured data to inform their analysis, Fisher added.

A graphic displays some of the key differences between structured and unstructured data.
The differences between structured data and unstructured data.

Beyond inheriting generative AI and unstructured data integration capabilities, Qlik's acquisition of Kyndi adds expertise.

Kyndi founder and CEO Ryan Welsh and other Kyndi personnel are now part of Qlik. The Kyndi team members bring with them knowledge about supporting unstructured data and developing generative AI capabilities not previously part of the Qlik platform. Adding that expertise is perhaps as significant as adding new capabilities, according to Fisher.

"The strategic integration of Kyndi's AI experts is as crucial as the technological acquisition itself," he said. "The team's expertise in AI significantly enhances Qlik's analytical and R&D capabilities [and] fortifies our product development trajectory."

Similarly, Leone noted that the acquisition of talent is just as important as the acquisition of technology.

"It's really a combination of the two, and that's why the acquisition makes so much sense for Qlik," he said. "They're gaining generative AI and unstructured data talent from the small team at Kyndi, but also their [intellectual property] that's deeply rooted in [natural language processing] and generative AI."

Future plans

Qlik did not reveal a timeline for when it expects to fully integrate Kyndi's capabilities into the Qlik portfolio.

Fisher, however, said Qlik plans to add Kyndi's technology to its generative AI roadmap and expects to introduce new capabilities resulting from the acquisition of Kyndi when it hosts Qlik Connect, the vendor's annual user conference, in early June.

Integrating the different technologies of two vendors following a merger or acquisition can sometimes be a difficult endeavor, but Qlik has been an aggressive acquirer over the past six years. It has proven adept at integrating its homegrown tools with those of both small vendors such as and CrunchBot, as well as larger ones such as Attunity and Talend.

"A pivotal moment is expected at Qlik Connect ... where we plan to reveal new products that significantly leverage Kyndi's [intellectual property]," Fisher said. "This event is a cornerstone in our journey to ... embed Kyndi's technology into Qlik's solutions, marking a new era of analysis and insight generation."

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.

Dig Deeper on Business intelligence technology

Data Management
Content Management