Welcome to the AI Agent & Copilot Podcast, analyzing the latest AI Copilot and agent developments from Microsoft and its partners, delving into customer use cases, and exploring how AI plus the Cloud helps customers reimagine their business.
In this episode, Will Hawkins, AI expert and founder of RitewAI, discusses early feedback, perceptions, and opportunities he’s identified with Model Context Protocol and Microsoft’s support for MCP.
Highlights
MCP Background and Use Cases (01:25)
Hawkins explains MCP (Model Context Protocol) as a universal connector between AI models and various data sources. MCP allows AI models to retrieve data, perform actions, and build robust workflows. He compares MCP to a USB-C connector for AI, enabling seamless data access and actions.
He describes how MCP can initiate workflows in other systems from an AI agent and provides an example of using MCP to get traffic updates and change routes based on the data retrieved. MCP creates a connection between the MCP server and the MCP client, allowing for multiple requests and actions. Will explains that MCP can perform functions preset by the MCP server, such as changing routes based on traffic data.
Implementation and Support for MCP (04:02)
He discusses the work required to use MCP in an organization, including enabling back-end systems to support MCP and mentions existing MCP connections for various data platforms like GitHub, Google Drive, Slack, and Postgres. He explains the effort to deploy MCP can be done locally or remotely, depending on the infrastructure and data sources.
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
Customer Interest, Security Concerns (06:02)
He notes that customers are showing interest in MCP, seeing it as a natural progression for data platform companies. There were recent updates to authentication methods. He suggests experimenting with MCP for low-risk data sets to build momentum and learn lessons.
While there are known vulnerabilities, developers can defend against them.
Microsoft’s Support for MCP (08:34)
Microsoft’s support for MCP is in Copilot Studio, Azure AE, and GitHub Copilot. In fact, he used MCP in GitHub to solve a coding issue. “I was actually using GitHub Copilot to work on some code, and I got an error message from a REST API that there was no documentation on. So I copied the error message, hit the Model Context Protocol on GitHub copilot, and asked it to search a web page. It found it for me and answered it and solved the the coding issue on the spot. So it is incredibly useful, and I think it’s great that Microsoft has adopted it.”
Meta and Apple, by contrast, are more cautious about MCP.
Vendor Support, Partner Opportunities (11:19)
He highlights the potential for MCP to create new business opportunities, such as selling data as a service.He mentions Zapier’s support for MCP and its potential to become an ISO standard. MCP is complementary to other technologies like Google’s Agent to Agent. He sees MCP as a universal format that connects any data source to any AI agent, with potential support from major vendors.
He identifies opportunities for Microsoft partners to advise customers and develop custom solutions using MCP. He mentions the potential for ISVs and partners to become MCP-compatible data platforms. Will discusses the importance of data quality and the competency gap in designing tools for AI agents.
Ask Cloud Wars AI Agent about this analysis
