Model Context Protocol, Agents and Semantic Layers

Exciting changes are underway in the world of data and AI. I recently interviewed Artyom Keydunov, CEO and co-founder of Cube, and we have discussed the Model Context Protocol (MCP) and its potential to reshape how AI agents interact with business systems. MCP is quickly becoming a new standard, allowing agents to not only access data across tools like Snowflake and Tableau but also to automate complex workflows and metadata management. This shift from isolated, point-to-point integrations to a composable, protocol-driven approach promises greater agility and scalability for data teams. Cube is at the forefront, building a platform where AI agents can act as tireless teammates-maintaining semantic models, propagating changes, and supporting data exploration 24/7. While people remain central for creativity and strategic choices, the future of data management will be increasingly driven by AI-powered automation, bringing us closer to the vision of a true data mesh. The pace of innovation is remarkable. We are eager to see how MCP and AI agents will continue to transform the data landscape.