Microsoft is bringing an A.I. chatbot to data analysis

Daily News
5 Min Read

Scott Guthrie, executive vice president of Microsoft’s cloud and artificial intelligence group, speaks at the Microsoft CIO Summit in Redmond, Washington, on Feb. 1, 2023.

Microsoft

Microsoft‘s hefty investment in artificial intelligence chatbots is finding its way into data analysis.

Following its reported $13 billion investment in OpenAI and its early efforts to embed the AI technology into its Bing search engine and other products like email, Word and Excel, Microsoft is rolling out a chatbot — or Copilot — that will allow users to make sense of information stored in corporate databases

related investing news

Broadcom is the 'most underappreciated AI beneficiary' after Apple deal, Bank of America says

CNBC Pro

Beyond the chatbot, Microsoft is launching a new brand called Fabric, which will bring together seven data products under one umbrella. It’s similar to the introduction of the Office productivity software suite in 1990, and Microsoft is touting cost-saving opportunities at a time when clients are tightening their belts.

Fabric is designed so that a customer can store a single copy of data and work with it in several programs. For example, the data can be tapped in Synapse Data Science for collaborating on AI models and Power BI business intelligence software for building charts and dashboards.

“You have one meter, so that if your BI system is underused, you can spend it on other parts, or if your data warehouse is underused, you can spend it on your BI meter,” said Scott Guthrie, executive vice president of Microsoft’s cloud and AI group. “You’re not having to pay for everything separately. And I think that ultimately yields some pretty significant cost savings to customers.”

A single Copilot will be available for the tools in the Microsoft Fabric portfolio, which includes Data Factory, Synapse Data Engineering, Synapse Data Warehousing, Synapse Real-Time Analytics and a new monitoring tool called Data Activator, in addition to Power BI and Synapse Data Science, a Microsoft spokesperson said.

The new technology doesn’t require technical expertise to operate. Anyone can open Power BI and type an idea for a report into Copilot, or click on a handful of automatically generated ideas. After a screen full of adjustable charts pops up, the user can type in a question about the data and receive an answer in plain English.

Earlier this month, Salesforce announced Tableau GPT, which will be able to make charts in response to text prompts.

Arun Ulagaratchagan, corporate vice president for Azure data at Microsoft, said Fabric is “more comprehensive than what has been announced in the market so far.” He said that Copilot will be able to write formulas using Microsoft’s Data Analysis Expressions language in Power BI.

Data for Fabric will sit inside a single storage system called OneLake, which users will be able to access as local files on PCs, similar to the way Microsoft’s OneDrive file syncing and sharing service works for Microsoft 365 productivity software subscribers. Users can view and take action on data stored in  Amazon Web Services’ S3 and Google Cloud Platform’s Google Cloud Storage, and they won’t need to make copies.

Boris Evelson, an analyst at technology researcher Forrester, said that based on how Microsoft has built the technology architecture, it “definitely will help them get bigger market share.”

Ulagaratchagan said Microsoft started working on Fabric two years ago. The company hasn’t yet finalized pricing.

Also on Tuesday, Microsoft said that in June it will start a preview for a Windows Copilot in its Windows 11 operating system.

“Just like you would with Bing Chat, you can ask Windows Copilot a range of questions from simple to complex,” Panos Panay, Microsoft’s chief product officer, wrote in a blog post.

WATCH: Microsoft CEO Satya Nadella on OpenAI relationship, generative A.I., Microsoft-Activision deal

Microsoft CEO Satya Nadella on OpenAI relationship, generative A.I., Microsoft-Activision deal

Share this Article
Leave a comment