It’s been quite a few years in the making. The latest version of Microsoft Dynamics 365 Contact Center (D365 Contact Center) is steeped in the systems deployed to support Microsoft’s own customer service organization. During its 49 year existence, the Colossus of Redmond had engineered a support system for customer care agents that was comprised of “16 different systems and over 500 individual tools”. The Azure Cloud based services that will be made available on July 1 marks what Jeff Comstock calls (in a blog post) “the latest milestone in our journey towards modernizing customer service”. It is a journey that it would like its entire base of enterprise software users to follow.

Like all things GenAI, the journey has accelerated significantly in the past three years, starting with the acquisition of speech processing and IVR-leader Nuance Communications in 2021 and the subsequent announcement of the first rendition of the Microsoft Digital Contact Center Platform in July 2022. Then ChatGPT happened (November 2022), making it possible for mere humans (including contact center agents) to converse with OpenAI’s GPT-x line of large language models. Three months later (January 2023), Microsoft increased its stake in OpenAI by more than $10 billion. The race was on to introduce enterprise hardened, GenAI-informed “copilots”, most notably Bing Chat for Enterprise and Microsoft 365 Copilot, which were introduced at its partner event, Microsoft Inspire 2023, with general availability expected in Q1 2024.

By February 2024, two enhanced versions of Microsoft 365 Copilot were made generally available and their prices were formalized. Microsoft Copilot for Sales, which connects with a company’s CRM system to extract insights and accelerate sales, and Microsoft Copilot for Service, which employs insights derived from a broad array of data sources to spark AI-powered conversations between contact center agents and customers. The big message in today’s announcement is the positioning of D365 Contact Center as a “copilot-first” offering. That makes Azure GPT, the version of OpenAI’s foundational LLM that runs in Microsoft Azure, the resource that powers both self-service and agent-assisted CX.

Copilot Studio is the Preferred Bot Development Toolkit

D365 Contact Center’s announcement lists “next-generation self-service” first among “key capabilities.” In doing so it cites the combination of “the best interactive voice response (IVR) technology from Nuance and Microsoft Copilot Studio’s no-code/low-code designer.” Copilot Studio strongly resembles the Power Virtual Agent Tool, introduced to “democratize conversational bots” in late 2019. Both rely, to a large degree, on a graphical interface to build intelligent virtual assistants. A major difference is that Copilot Studio has evolved into a platform that responds to natural language input and powers resources that can take actions “as an independent agent” when it perceives the need. For instance when an email arrives from a customer, including communications and transaction histories and it can retrieve relevant information from back-office systems to respond to queries.

In a demo of Copilot’s capability, product t manager Alan Ross showed how a copilot responds when a customer asks an agent “how much does it cost to ship my package”. It responded to that query by asking one of its own: “What is your Zip Code?”. In a process that is essentially filling out a Webform, it carries out a conversation in natural language. Principal Program Manager Marcus Schmidt (an Intrado veteran) added that the Copilot would respond differently when a customer is authenticated. In this case it would start the conversation by asking something like “will we be shipping this to your home address?” Their point is that the answers are dynamically generated based on Natural Language Understanding, intent matching. They called it “putting a wrapper around an API” that automatically prompts the agent or asks the customer directly to provide their Zip Code. Then it consults the proper system of record to generate natural language result or engage in a conversation that solicits more required info from the customer.

A “Hybrid Approach” to Migration from Existing Solutions

GenAI, not Microsoft Dynamics, is core for self-service, agent assistance, and operational efficiencies. Based on its own customer service challenges, Microsoft understands viscerally that the first-order challenge for CX professionals is overcoming the chaos and extra expense caused by “operating with a patchwork of disparate systems.” Going “Copilot-first” is a significant repositioning of D365CC. For one thing, it recognizes that the information that responds to customer or agent queries often come from a “shadow system of record” rather than Microsoft-branded sources like Sharepoint or Dynamics 365. D365 Contact Center is already “open” to integrating with Salelsforce’s cloud-based resources and promises more connectors, in response to market demand.

Looking specifically at the self-service side, while it notes that its solution combines “the best of interactive voice response (IVR) technology from Nuance and Microsoft Copilot Studio’s no-code/low-code designer”, its announcement sheds little light on strategies for migrating or replacing service creation platforms, meaning Nuance Mix. When asked whether it has a migration strategy for existing use cases, the answer is a qualified “yes.” Alan Ross noted that enterprise customers will start with one-to-one migration of a selected use case to the new environment. In the course of what could be a labor-intensive process, they are likely to determine that “putting a wrapper on existing APIs” using a no-code tool, is preferable the one-off approach. The path will be from Mix to Copilot Studio to a hybrid that leans more toward GenAI.

Late to the Party? That’s For the Enterprise Market to Decide

The new D365 Contact Center will be generally available July 1. In the mean time, in addition to its own customer support operations, there are testimonials from 1-800-Flowers, Mediterranean Shipping Company and Synoptic who have experience with the service. Each recognizes the value of GenAI, embedded in Microsoft’s cloud, to support improved customer experience and employee productivity. Yet, every player, large and small, in the Contact Center as Service (CCaaS) marketplace has been touting their approach to GenAI spanning both CCaaS and UCaaS.

Microsoft is correct to make GenAI the core of a differentiated CCaaS offering. It is a newcomer to CCaaS, yet it has spent more than a decade establishing its bonafides in Conversational AI (the precursor to GenAI). It formally launched Microsoft Cognitive Services in 2016, and first introduced Power Virtual Agents in 2019. So, through the lens of a copilot-led offering, Microsoft can be seen as a leader. Still, it is already taking on formidable competition, as will be made evident at Customer Contact Week (CCW), where it is holding its coming out party. Attendees will have an opportunity to conduct their own investigations into the comparative merits of Microsoft’s Copilot-first approach versus the likes of Amazon Web Services, Zoom Contact Center, Five9, Verint, Zendesk, UJET, Dialpad, 8×8, and others that I am forgetting. All of them seek to differentiate by emphasizing the core technology and tooling they provide to introduce GenAI into customer and agent workflows. Many of them (though it may be pale) also have more customer testimonials and use cases to showcase.

Categories: Intelligent Assistants