Power Virtual Agents and Copilot — Your Agent of Change

We’ve been working our way through the introduction of Microsoft Copilot into Power Platform, highlighting the features and functionalities this generative AI tool brings to each application within the platform, including Power Automate, Power Apps, Power Pages, Power BI, and Power Virtual Agents. Here, we’ll showcase how Copilot in Power Virtual Agents may be the change agent you’ve been waiting for.

Virtual Agents Are Already Smart, Aren’t They?

You may be thinking, Power Virtual Agents already leverages AI to work its magic — so what’s new here? Sophisticated chatbots like Power Virtual Agents indeed leverage artificial intelligence to locate and deliver the assistance customers need. Still, they require some education in order to get as smart as they are. And that education typically comes in the form of a human building and curating dialogs step by step, which takes time. In fact, it’s not uncommon for chatbot builders to spend weeks or months developing and refining their bots.

When building a bot, developers typically need to create multiple topics, each addressing a specific issue or task. This process can be time-consuming and often demands specialized knowledge in conversational user experiences and technical domains. However, this isn’t always feasible, especially in organizations with diverse teams collaborating directly in the Power Virtual Agents authoring canvas.

How Copilot Changes the Dialog

Introducing Copilot in the Power Virtual Agents preview canvas.* Copilot simplifies the topic creation process by allowing developers to describe their objectives in plain language — instantly creating a helpful chatbot. With no manual authoring, an “empty” chatbot can respond to your users with relevant information, even if you haven’t manually created a topic for it.

Simply describe what you want your topic to do, and Power Virtual Agents creates it for you, complete with conversational responses and multiple types of nodes. Use the suggested topic as is or as a starting point for further iteration. Copilot leverages the natural language understanding models within Power Virtual Agents and Azure OpenAI to:

• Understand the bot maker’s intentions by parsing their request.
• Apply knowledge of how nodes within a topic should work together and be structured for optimal results.
• Generate a sequence of interconnected nodes to form a complete topic.
• Use user-friendly language in any node containing customer-facing text aligned with the bot maker’s request.

As a result, Copilot for Virtual Agents transforms tasks that once took hours into a matter of minutes.

Let’s Discuss What’s Next

We think Copilot for Power Virtual Agents is a significant advancement that simplifies your development team’s job and enhances your users’ customer experience, and it’s possibilities for the future are enormous. Below are some potential ways Copilot can change things going forward.

Enhanced Customer Support

Copilot could assist your virtual agent in handling customer queries more efficiently. It can analyze and interpret customer messages, extract relevant information, and suggest appropriate responses. For example, when a customer asks about order status, Copilot can extract the order number, check the database, and provide the latest shipping information — all within seconds.

Personalized Sales Recommendations

Copilot could use its natural language processing capabilities to understand customer preferences and recommend products or services that match their needs. This level of personalization can boost cross-selling and upselling efforts. For example, if a customer asks for recommendations for a new laptop, Copilot can consider their budget, usage, and brand preferences to suggest the ideal options from your product catalog.

Seamless Handoffs to Live Agents

Copilot could identify when a customer issue requires human intervention and seamlessly transition the conversation to a live agent. The handoff ensures that complex issues are handled by the right experts while maintaining a smooth customer experience. For example, suppose a customer has a technical problem that the virtual agent can’t resolve. In that case, Copilot can connect them to a live support agent, providing the agent with a summary of the issue for a seamless transition.

Language Support and Translation

Copilot’s language capabilities extend beyond English. It can assist in real-time language translation, allowing your virtual agent to communicate with customers worldwide breaking down language barriers. For example, if a customer messages in Spanish, Copilot can translate their query into English for the virtual agent to understand and respond appropriately.

Content Generation and Updates

Copilot could help keep your knowledge base up to date by generating responses and updating FAQs based on the latest information. For example, when you launch a new product, Copilot can automatically generate answers to frequently asked questions about the product and update the knowledge base.

Something to Talk About

Copilot takes Microsoft Power Virtual Agents to new heights of effectiveness and customer engagement. By combining the power of Copilot with the already significant capabilities of Power Virtual Agents, your organization can quickly deploy virtual agents to provide personalized, efficient, and empathetic customer experiences, setting your business apart while optimizing your internal resources. Velosio is here to help you pilot your Copilot journey. We’ve got tools, resources, and expertise ready. Contact us and we’ll get you the information you need.

*Copilot for Microsoft Power Virtual Agents is currently in private preview mode. Keep an eye out for updates from us, as this is one to chat about.