Migration from Power Virtual Agents to Copilot Studio: What You Need to Know

Microsoft’s low-code/no-code chatbot creation platform, Power Virtual Agents (PVA), is currently referred to as Microsoft Copilot Studio. Without the need for sophisticated programming knowledge, it enables you to build intelligent chatbots (also known as copilots or Agents) that can communicate with users across various platforms, including websites, mobile apps and Microsoft Teams. Microsoft also offers an Applied Skill course on Agent creation named PL-7008: Create Custom Agents in Microsoft Copilot Studio, where one can seek help through a guided project, practice skills and learn how to design custom agents in Copilot.

Power Virtual Agents (PVA)

Earlier with Power Virtual Agents, one was able to create conversation flows with a no-code, graphical user interface, connect with Power Automate to initiate processes (such as data updates or sending emails), establish a connection with data sources like APIs, Dataverse and SharePoint, and also post bots to websites, Teams and other platforms. Analytics dashboards could be used to track and evaluate chatbot performance.

Rebranding PVA to Microsoft Copilot Studio

To align Power Virtual Agents and Copilot Studio with the Microsoft Copilot ecosystem, Microsoft changed the names in late 2023. Now, Copilot Studio (Fig. 1):

• Creates Agents with AI capabilities, which are more sophisticated than conventional chatbots/copilots.
• Features seamless integration with both custom plugins and actions, as well as Microsoft 365 Copilot.
• Allows natural language with the capabilities of Gen AI (orchestration) – one can simply type commands/prompts to the Agents and it will create the flow.
• Enables autonomous agent capabilities, allowing bots to operate and make decisions independently, providing more intelligent and contextual answers through the use of Generative AI (OpenAI models and Azure AI).

Fig. 1: Microsoft Copilot Studio home page

Transition to Microsoft Copilot Studio

Not merely a rebranding, Microsoft’s move from Power Virtual Agents (PVA) to Microsoft Copilot Studio was a deliberate step to better align with the company’s new AI-first, Agent-driven vision for the Power Platform and Microsoft 365.

1. Aligning strategically with the “Copilot/Agent” vision

Word, Excel, Power Apps, Dynamics 365 and other Microsoft products are all being consolidated into the Copilot ecosystem. Initially designed as a bot platform, Power Virtual Agents’ function grew as AI copilots/Agents became essential to Microsoft’s product strategy. The objective is to consolidate all AI conversational experiences (bots, assistants and copilots/Agents) under Copilot Studio so that users may create a single type of AI assistant rather than distinct tools for various platforms.

2. From Scripted Chatbots to Generative AI Copilots

Power Virtual Agents were organized Q&A-style bots that operated on predetermined subjects and flows. However, consumers were calling for AI that could understand open-ended inquiries and generate intelligent responses. The addition of Copilot Studio introduced dialogue with Generative AI (GPT models) (Fig. 2), enabling the system to act and reason rather than react, thereby facilitating dynamic interaction with business information.

Fig. 2: Gen AI Orchestration in Agent overview

3. Integrated Development across the Power Platform

Microsoft aimed to facilitate the integration of chat interactions with Dataverse, Power Automate and Power Apps. Copilot Studio now allows you to:

• Create plugins or actions similar to Power Automate flows by adding tools (see Fig. 3).
• Use Microsoft Graph connectors to interact with Microsoft 365 data.
• Leverage context memory to build multi-turn conversations using prompts.

With this unification, Agents are no longer standalone bots but are now components within the broader Power Platform solution.

Fig. 3. Letting Agent add tools

4. Easy and Smart Authoring Experience

Formerly in PVA, topics, conditions and trigger phrases had to be created manually. Now in Copilot Studio, you can use natural language to express what you desire to see. Topics and flows are automatically generated by the system with Gen AI orchestration and can be graphically edited and improved. By lowering the entry barrier with GPT models introduced in the Copilot studio, business customers and citizen developers may produce agents more quickly.

5. Better Governance, Security and Integration

Copilot Studio is built on Microsoft’s AI safety and compliance framework, integrated with Azure OpenAI Service and Microsoft Purview governance. Organizations can achieve centralised governance for all Agents, consistent security and compliance controls and audit and monitoring tools integrated into the Microsoft ecosystem.

Future of Microsoft Copilot Studio

By combining conversational AI technologies like Power Virtual Agents into a single, more potent platform, Microsoft is rebranding to Copilot Studio as part of its strategy to streamline and simplify its AI product offerings. This change aims to clarify any misunderstandings regarding the term “Copilot,” improve differentiation between various Copilot products and explicitly match client investments with the best solutions for their specific requirements. The platform’s development to incorporate generative AI, multi-channel deployment and deep integration throughout the Microsoft ecosystem is also reflected in the name change. In short, Microsoft migrated Power Virtual Agents to Copilot Studio to unify conversational AI, integrate generative AI, simplify development and align with the enterprise Copilot vision.

About CloudThat