Microsoft is conducting its "AI Roadshow" in 60 cities worldwide. These one-day events aim to bridge the gap, allowing business professionals interested in AI to engage directly with Microsoft engineers and sales teams to discuss how to build "AI-first" enterprises and the prospects for the application of their core technology, Copilot.
At today's London event, Microsoft announced new features for Copilot Studio (formerly known as Power Virtual Agents), which will help businesses build and deploy AI agents more conveniently. Microsoft has metaphorically described AI agents as "the new applications of the AI era," emphasizing their ability to provide intelligent solutions for specific business problems.
The highlight of the new Copilot Studio is its "building block" development approach. Operations managers can quickly build AI applications without a deep programming background, simply by configuring rules, setting trigger conditions, and describing workflows. Users can start with pre-built agent templates and adjust them according to their business operational needs, including modifying behaviors, adjusting data inputs, and customizing prompts and vocabulary used by the agent.
This update includes four important new features:
- Launch of ten pre-built autonomous agents for sales, service, finance, and supply chain
- Introduction of an automatic trigger mechanism based on business signals
- Support for dynamic task planning, enabling workflow adjustments based on context
- New activity tracking feature for easier human oversight and debugging
The case of UK pet supplies retailer Pets at Home demonstrates the practical application value of Copilot Studio. The company has automated its profit protection workflow using AI agents, helping its over 450 stores effectively prevent fraud and theft. The company's CIO, William Hewish, stated that the integration of AI allows the team to focus more on analytical work that requires professional judgment.
Management consulting firm McKinsey is also actively embracing AI transformation, using Copilot to build an intelligent system for handling new client proposals, achieving rapid expert team matching, and other functions. McKinsey's senior partner, Rodney Zemmel, expressed great expectations for the future development of AI agents.
It is worth noting that Microsoft has particularly strengthened its security control mechanisms in this update, allowing businesses to set access controls, data protection policies, and other security guardrails to ensure that AI applications operate within a controlled scope. This reflects Microsoft's commitment to innovation in AI while seriously considering how to ensure that technology development progresses in a positive and predictable direction.