This could be an early glimpse of the future integration of AI into enterprises. MicroStrategy On Tuesday, the company announced an addition to its platform which simplifies the access to data for business analytics within organizations.
MicroStrategy Auto, a customizable AI robot that can be customized by the user, is said to offer a faster and simpler way for anyone to receive business intelligence within an organization. Auto is the latest enhancement of MicroStrategy AI. It was released in October 2023 and allows users to quickly build AI applications using trusted data.
The company said that Auto can be used as a standalone application or embedded in third-party apps, and it offers full customization. Its appearance can be customized, as well as its language style and level.
Auto is powered by AI generative, so users can communicate with it using natural language.
“We use GPT4 for the backend — for figuring out what the user is asking for and how to answer the question,” explained MicroStrategy Executive Vice President and Chief Product Officer Saurabh Abhyankar.
He told TechNewsWorld that the difference between MicroStrategy’s large language model and the general purpose LLM is the addition of an analytic database structure. “If you ask me how many hats are there at store X the LLM determines what the user wants, while the MicroStrategy Layer executes the query and brings back the data, as well as applying security and rules in order to calculate inventory.”
He added, “You’ll need both in an enterprise analytics scenario, because a ChatGPT chatbot doesn’t have context, business expertise, security and governance to answer such a question.”
Unlocking the User Value
Auto’s AI can enable users to make faster and more effective decisions. It makes applications smarter, and puts enterprise analytics at their fingertips, no matter how skilled they are or what application they use.
Business intelligence can be easily integrated into business decisions by using simple language.
We believe that MicroStrategy’s AI will unlock a lot of value for a wide range of users by providing them with deeper insights, which previously required more clicks or more granularity in order to understand. It’s great for user self-service,” Nena Piedskalny said in a Federated Co-operatives Limited statement.
Mark N. Vena added that “Giving employees more access to data on business intelligence can benefit a firm by fostering informed decisions across departments, enabling flexibility in responding to changes in the market, and promoting culture of data-driven decisions.” SmartTech Research San Jose is located in California.
“However,” he said, “easier access to data on business intelligence may lead to potential harms, such as data breaches and misuse of sensitive data, or compromising competitive edge if properly managed and secure.”
Rob Enderle is the president and principal analyst of the Enderle Group“They’re also able to run locally because they use smaller libraries.” They can also run locally, because they use smaller library.
Enderle also added that custom enterprise bots are safer than general purpose bots. He said that these bots are usually derivatives from large LLMs. However, because they’re smaller and more focused in theory, there is less chance they will do anything you don’t want.
AI: What You Need to Know
Businesses can address their concerns over data sharing by creating custom generative AI bots. Will Duffield said that there is always some anxiety when it comes to giving up proprietary data to a tool which will then iterate the information and possibly re-present the data in another way. Cato InstituteA Washington, D.C.-based think tank.
TechNewsWorld reported that consumer-centric bots allow firms to use conversations from consumers to improve bots. “That would not be the case for a lot these business tools, because the way the information can used will be contractually stipulated.”
Abhyankar continued, “Enterprises are reluctant to give all of their data to LLMs that have a broad range of applications.” They do not want to use their data for training the LLM because they are concerned about the data being leaked.
MicroStrategy is storing data within the customer environment, explained he. The LLM doesn’t get trained with the data, but only a few bits of meta-data are sent. “We are able to do this because MicroStrategy does the calculations. The LLM is not required to perform these calculations so it doesn’t require all the data,” explained he.
The LLM is also prevented from having hallucinations for the same reason. Abhyankar stated that “LLMs are probabilistic by nature.” You can ask the LLM questions, but it can give you different answers to the same question. It’s not ideal in a business setting.
We can avoid the probabilistic problem by using the MicroStrategy layer to run calculations based on business logic that our customers have encoded into our platform.
He added, “The challenges of data-sharing and hallucinations have been largely eliminated because we only use the LLM for cognitive skills and we use customer data in MicroStrategy in a trustworthy fashion.”
Pumping up Productivity
It can be beneficial to productivity for enterprise staff to have more access to business intelligence. Enderle stated that business intelligence should enable decision makers to take better and timely decisions. This will lead to greater operational success.
MicroStrategy Auto should increase productivity, especially for data analysts. Abhyankar stated that it makes data analysts more efficient because they can accomplish more in less time. It is a productivity boost.”
The analyst can now focus on more important tasks because there are fewer questions or requests coming from the users.
Sharad Varshney is the CEO of OvalEdgeData analytics is being impacted by generative AI technology across the board, according to, an Alpharetta-based data governance consultant and provider of end-to-end catalog solutions. “They simplify the data discovery process, making it easier for teams like HR or marketing that aren’t analytics-focused to make use of company data assets,” he said.
He said that “however,” the data must be accurately governed. Although a generative AI can quickly find data and contextualize it, it does not account for data lineage, quality or access.
He continued: “Once the data is found, it’s important to have policies in place so that the user who requests the data can extract the data with the appropriate access rights.” Then it must be subjected to various quality measures for duplication, consistency, and other factors, before being classified or cataloged. Only then can it be analyzed.”
He added that “fortunately” there are tools that automate these processes of governance and others, making data analysis and visualisation very simple.
