New embeddable MicroStrategy Auto bot expands AI beyond BI

New capabilities

Generative AI has been the biggest trend in data management and analytics for more than a year, ever since OpenAI's release of ChatGPT in November 2022 marked a significant improvement in large language model (LLM) capabilities.

Two primary reasons generative AI has been so ubiquitous are its potential to both enable non-technical users to work with data as well as help data experts be more efficient.

By developing tools built through integrations with LLMs such as ChatGPT and Google Gemini, analytics vendors can provide NLP tools to interact with data, rather than code.

Many vendors developed their own NLP capabilities in recent years, but they were narrow in scope due to the tool's limited vocabularies. They still required users to phrase queries in an exact manner and delivered responses in language that necessitated data literacy training.

LLMs, unlike the NLP capabilities developed by analytics vendors, are trained on public data and have vocabularies as extensive as a dictionary. That enables users to phrase queries and other prompts in true natural language, which reduces at least some need for data literacy training and enables more non-technical workers to use analytics in their workflow.

In addition, LLMs can be trained to translate text to code as well as generate code on their own. That can save developers and trained data analysts from writing the code required to develop and update data products such as dashboards and reports.

MicroStrategy aims to use generative AI to open analytics to non-technical users as well as make trained experts more efficient, according to the vendor.

Auto, which is available at no extra cost beyond what customers already pay for their MicroStrategy AI, extends the reach of  Microstrategy AI beyond the BI environment.

With Auto embedded into other applications, users don't have to leave their normal work environment and enter MicroStrategy's to look at dashboards and read reports to derive insights from data. While still using third-party applications such as Salesforce, they can ask questions of data using natural language and receive responses also in natural language, then take action based on the insights derived from their queries.

The MicroStrategy AI bot joins Auto SQL so users can automate SQL code generation, Auto Dashboard to enable development of dashboards using conversational language and Auto Answers for support when working with data.

NLP-based SQL generation, dashboard development and support are not unique among analytics vendors, Henschen noted. Neither is the use of MicroStrategy's analytics engine and semantic layer to improve the accuracy of natural language prompts and responses.

However, Auto as a standalone application does represent evolution. So does the speed with which MicroStrategy is making its generative AI capabilities generally available, according to Henschen.

While many data management and analytics vendors unveiled generative AI capabilities earlier than MicroStrategy, the vendor was among the first to make such capabilities generally available.

ThoughtSpot revealed Sage, a tool that enables true NLP, in March 2023 while Tableau unveiled Pulse and other generative AI tools in May 2023.

MicroStrategy AI was introduced in October and immediately became generally available whereas many of the tools introduced by competitors are either still in preview or were only made generally available after a lengthy preview process.

Responses from users following MicroStrategy AI's release led to updates in the vendor's December and March platform releases, including attempts to address the accuracy of generative AI outputs by using vectors in concert with existing MicroStrategy capabilities.

"The performance and accuracy improvements that we made in the December release and the March release all came from the feedback we got," said Saurabh Abhyankar, MicroStrategy's chief product officer. "All of it is to help make [MicroStrategy AI] better at answering questions."

Auto resulted in part from feedback following the release of MicroStrategy AI, he continued.

Companies wanted to provide generative AI capabilities to more users, but they were limited by MicroStrategy's environment. They didn't ask specifically for a bot that could be embedded into other applications, but their questions provided MicroStrategy some of the impetus for the idea of an embeddable AI bot.

A second motivator for Auto's development was an attempt to stand apart from other analytics vendors, according to Abhyankar.

"Now, everyone is doing AI, so we asked what we could do that is unique and different, [which led to] this notion of getting beyond the core analyst audience to the regular person," Abhyankar said.

Next steps

With Auto now part of MicroStrategy AI, one aspect of the vendor's roadmap is to expand the portfolio with additional generative AI tools, according to Abhyankar.

Auto SQL enables code generation, Auto Dashboard is aimed at helping users quickly build and update dashboards and Auto Answers can help anyone in need of assistance while using MicroStrategy's tools.

The ability to use Auto as a freestanding app … will take analytics beyond the confines of analytics/BI interfaces.

Developing data models is another area where generative AI can help users be more efficient, Abhyankar noted.

"AI is an opportunity to make every persona in the BI world more productive," he said. "Auto SQL is the first step for that data persona. What about generating data models? We have a semantic layer and data modeling, so what about powering that with AI? We think there's a huge opportunity there."

In addition, MicroStrategy's product development plans include finding new ways of getting data to workers on the edge, Abhyankar continued.

The vendor was an aggressive developer of its mobile app at a time when many BI vendors struggled to build effective mobile tools. In addition, it was an early pioneer of embedded analytics and has now built a bot that can be embedded in end users' workflows. Finding ways to enable MicroStrategy's use outside the MicroStrategy environment will continue to be a focus.

Henschen, meanwhile, said that the vendor would be wise to integrate with a wide variety of LLMs so that customers can use the model of their choice when developing generative AI applications.

MicroStrategy has an integration with OpenAI through Microsoft and has said it plans to add support for Amazon Bedrock and Google Palm 2, he noted. However, neither has been formally launched.

"I'd like to hear the update on [support for Amazon Bedrock and Google Palm2]," Henschen said. "Having a choice, including smaller, more focused and more affordable models, would complement MicroStrategy's very strong multi-cloud capabilities."

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.