AI Studio highlights new Alteryx Generative AI offerings

Alteryx looks to build on the momentum in the Generative AI space it built earlier this year when it was in the vanguard of actually getting product into general availability.

Suresh Vittal, chief product officer at Alteryx

Today, analytics cloud platform vendor Alteryx is announcing three new offerings, including a generative-AI based deployment-agnostic AI Studio designed for no-code users – an industry first.

Alteryx announced two Generative AI products being released into general availability at their Inspire event in Las Vegas last spring, which utilized their Alteryx AiDIN artificial intelligence, machine learning, and generative AI engine, which was also announced. At the time, the company considered this to be a major coup, because while everyone is jumping on the GenAI bandwagon, almost all at that time had products as demos and in beta only. Alteryx, on the other hand, had product in market. Suresh Vittal, chief product officer at Alteryx, said that optimism had been born out.

“The market has been trending extremely well us since the spring,” Vittal stated. “Having those two products, Magic Documents and Workflow Summary, in the market has proven to be a really good thing. It has allowed us to get feedback from customers.

Vittal said that the pickup rate of their existing install base on Magic, whicg uses Generative AI to create initial drafts of dynamic content, has been phenomenal.

“Since we announced Magic, more than 50% of our customer base has embraced the Magic features, in only four months,” he noted. “It lets them create in a different language, or use different tones on voice.”

Workflow Flexibility was designed to deal with change management, and has also had a positive customer response since its release.

“Workloads like code may not be well documented, or the analysts who originally wrote them might change,” Vittal said. “Change management was a real challenge. Workflow Flexibility created an automated layer of governance, Over 200 of our customers have begun to use it, because we made this capability available.

Looking to build on these successes, the new announcements begin with the Alteryx AI Studio. It gives organizations the power to select an LLM of their choice from a list of available options, then tune it using their own custom business data to adapt to the specific context of their business – thus enabling analytics automation without having to write code. It seamlessly integrates to Alteryx Designer so customers can easily consume their models through existing workflows and construct applications with a conversational interface.

“This is not a general purpose design studio,” Vittal noted. ‘We have been training multiple foundational models, and we believe we have a unique repository of analytics, which make being an analyst much more powerful and intuitive. LLM lets you train operators, but with the Alteryx Ai studio, customers can bring their own data and build on top of our model.”

Vittal indicated that Alteryx had trained open source models uniquely to Alteryx’s data sets.

“Then we created the Studio environment where customers build on top of our model with their own data,” he added.

Alteryx expects that larger enterprises will be especially attracted to AI Studio.

“For a lot of customers, the training we do will suffice, but for many more, likely about 50% of the market, and especially in the  large enterprise, they have been creating unique data assets themselves,” Vittal said. For additional training on top of that specific set of data, they need sophistication, data scientists and developers. Those are the customers that will care most about this.”

Another capability being announced is Playbooks, an upcoming Auto Insights feature. Users identify their role and company, then Playbooks will suggest relevant and important analytics use cases to explore and build.

Vittal explained why this is important.

“When customers start an analytics project, they are often stopped in the early cases by uncertainty over use cases, and how others have done them,” he said. “There are always lots of questions. For customers like that, we want to make it easier for them to jump start their mission. All the user has to do is add their job function, and with that we are able to help them get started.” He noted that the company had an earlier capability, BluePrints, which was similar, but that Playbooks enhances this even further.

Alteryx is also announcing Alteryx Marketplace, which is available now.

“We have an amazing community of end users and partners,” Vittal said. “We want to support that community even more. The Marketplace will be in place to discover creator-supported solutions where they can build IP on top of Alteryx.”

Alteryx will also soon be introducing chat interfaces into the analytics process to further lower barriers to entry for users.

“People should be able to design an analytics project from anywhere at any time, and we will make this easier with new Copilot and chat interfaces,” Vittal indicated. Collaboration across different analytics roles in organization has tended to be informal, and has been hamstrung as a result. We will address this through multi-model analytics.”

The multi-modal analytics and the chat and Copilot interfaces will be available soon.