According to Andrew Ng (@AndrewYNg), a new short course titled 'Build AI Apps with MCP Servers: Working with Box Files' has launched, developed in partnership with Box and taught by their CTO, BenAtBox. The course demonstrates how the Model Context Protocol (MCP) can eliminate the need for custom integration code by allowing AI applications to offload file operations to dedicated servers. Learners will gain practical skills in building LLM-powered document processing apps using the Box MCP server, leveraging tools that large language models (LLMs) can access directly. The curriculum also covers designing multi-agent systems with Google's Agent Development Kit (ADK) and orchestrating workflows via the Agent2Agent (A2A) protocol. This approach streamlines file access for AI applications, accelerating secure enterprise document processing and opening up new business opportunities in AI-driven workflow automation, as cited by Andrew Ng's official Twitter announcement.

According to @perplexity_ai, workspace admins can now leverage the Model Context Protocol (MCP) beta to build custom deep research connectors. This empowers organizations to integrate proprietary systems and various third-party apps, enabling teams to search, reason, and act on internal knowledge alongside web results and pre-built connectors (source: @perplexity_ai, 2024-06). Available to Team, Enterprise, Edu admins, and Pro users, this advancement opens new business opportunities for organizations seeking to enhance AI-driven knowledge management and workflow automation.