The "next generation" of AI product enhancements for Microsoft's business apps portfolio were unveiled today. They discuss both Dynamics 365, the company's package of enterprise resource planning (ERP) and customer relationship management (CRM) solutions, and Power Platform, Microsoft's collection of low-code tools for creating apps and workflows.
Charles Lamanna, CVP of business apps and platform at Microsoft, hailed the upgrades as the obvious next step in Microsoft's automation journey in an interview with TechCrunch. The new features follow the introduction of OpenAI text-generating AI models in Power Platform four years ago and the more recent introduction of generative AI capabilities in Viva Sales, Microsoft's seller experience app. Both are powered by technology from AI startup OpenAI and developed using the Azure OpenAI Service, a Microsoft service that offers enterprise-tailored access to OpenAI's API.
Lamanna noted that Microsoft has a long-standing partnership with OpenAI to commercialise the vendor's technology in Microsoft's own products and through the Azure OpenAI Service. "Over the last four years, we've been on a journey to bring generative AI and foundation models to the workplace," Lamanna said via email. And we've evolved to the stage where technology and products can give customers truly revolutionary experiences.
Microsoft is releasing a feature called Copilot for Dynamics 365, which, broadly speaking, intends to automate some of the most repetitive sales and customer service processes (borrowing branding from GitHub's Copilot service).
For instance, Copilot can assist in writing email responses to clients in Dynamics 365 Sales and Viva Sales, as well as in producing an email summary of a Teams meeting in Outlook. According to Lamanna, the meeting summary combines information from the recorded Conference conversation with information taken from the seller's CRM, such as product and pricing details.
At runtime, Lamanna continued, "We access data from customers' CRM, ERP, and other enterprise data sources safely and intelligently. We blend company data with underlying knowledge using massive language models to provide responses that are customised for each consumer. It's important to note that we don't train the models using consumer data.
In Dynamics 365 Customer Service, Copilot may create "contextual replies" to client questions submitted via chat or email and offer customer support representatives a "interactive chat experience" that incorporates case studies and knowledge sources. They go along with Power Virtual Agents, Microsoft's chatbot builder, which now has a "conversation enhancers" function that enables businesses to connect a bot to websites or knowledge bases in order to leverage that information to answer queries that the bot hasn't been trained to answer.
Conversely, conversation boosters support Microsoft's AI Builder tool's new "GPT" paradigm, which enables businesses to integrate text-generation capabilities into their Power Automate and Power Apps solutions. A marketing manager might use the GPT model to generate targeted content ideas by entering particular keywords or subjects, according to Lamanna, while a researcher could use it to summarise text from weekly issued studies and have it forwarded to their email.
One might be hesitant to develop an app employing Microsoft's technology given the company's recent push into generative language, such as Bing Talk, for fear of it going off the rails. But, according to Lamanna, each customer's CRM, ERP, and other data sources serve as "grounded in reality" for conversation boosters, the GPT model, and Copilot as well.
Users are urged to check the accuracy of AI-generated content before utilising it because it is always properly labelled. To help the user more easily confirm the accuracy of the response, we additionally indicate the sources from which the answer was gathered where appropriate, according to Lamanna. In case any problems bypass the above lines of defence, we have monitoring and controls in place to allow us to immediately respond with manual intervention.
Users are not required to check the accuracy of the content, but that doesn't stop them from doing so. Time will tell if it develops into a problem, but research on automation bias, or people's propensity to put too much faith in AI, suggests that it may.
Fortunately, the rest of Copilot's features are less likely to cause issues.
With Copilot in Dynamics 365 Customer Insights and Dynamics 365 Marketing, marketers can construct target segments by describing the segment in their own words and receive suggestions regarding client categories that they may not have previously considered. Users can also request specific themes from Copilot, which generates them by drawing from "a range" of online resources as well as an organization's current marketing emails, to gain ideas for email campaigns, according to Lamanna.
Microsoft is lagging behind in some areas. Salesforce, the CRM white elephant in the room, has been injecting (or at least attempting to inject) AI-powered capabilities into its CRM family of products for years. AI has also been adopted by startups like Glint, largely to streamline customer support processes. But, as more and more marketers indicate they intend to include AI into their content strategies, it may not necessarily matter who is first to market, but rather who implements it first at scale.
The mission-critical customer and company data sources CRM and ERP have long been, but Lamanna noted that they frequently necessitate time-consuming operations like manual data input, content creation, and note-taking. "Dynamics 365 Copilot automates these time-consuming procedures and frees up the workforce's full creative potential."
Copilot, a component of Microsoft's business management tool Dynamics 365 Business Central, aims to make the creation of e-commerce product listings more efficient. According to Lamanna, Copilot can generate descriptions for product characteristics like colour, material, and size that may be customised by changing elements like tone of voice, format, and duration.
Similar to Shopify's recently released AI-generated product descriptions service, Lamanna hinted at this similarity. He noted that Shopify users of Business Central can add products with AI-generated descriptions to their Shopify stores in "in a few clicks" (after hopefully checking them for accuracy).
In addition, Copilot in Microsoft Supply Chain Center, which is riding the trend of automation in the supply chain sector, may proactively identify challenges like weather, finances, and location that could affect supply chain procedures. After that, supply chain planners have the option of having Copilot automatically create an email to notify any affected partners.
Lamanna contends that even basic AI-enhanced procedures like automating emails can result in a discernible increase in productivity.
"According to a recent survey we conducted on business trends, 90% of employees believe AI would eliminate repetitive duties from their professions. AI-powered assistants are increasingly standard in commercial apps, according to Lamanna. "We think Dynamics 365 Copilot will help workers complete tasks more quickly so that businesses can devote more time to the imaginative, cutting-edge aspects of their work, like developing enduring relationships with customers."
The truth is obscured once again by some marketing speak. Nonetheless, it is evident that Microsoft is continuing to make investments in automation and artificial intelligence. Microsoft just increased its investment in OpenAI by billions of dollars, and the business is keen to see a return on its capital.
According to Microsoft, Copilot will be a free add-on to existing Dynamics 365 subscriptions like Dynamics 365 Sales Enterprise and Dynamics 365 Customer Service Enterprise. Beginning on March 6, it will go into preview. Afterwards, it will become generally available.
Follow Us