Everyone’s talking about MCP… but not everyone’s convinced.
As organizations explore how to connect AI agents with operational data, some see MCP as the next big standard for secure connectivity. Others argue it’s still too early — that agentic systems need better orchestration, context management, and human oversight before any single protocol can define the space.
In this episode of MCP [Un]Plugged, Jake Bengtson, VP of AI Solutions at Striim, sits down with Alexander Noonan, Developer Advocate at Dagster Labs, for a candid, forward-looking conversation on what MCP really is right now, and what it could become.
Attendees will learn:
- What MCP represents in the broader evolution of agentic AI
- How orchestration, governance, and connectivity intersect in the era of intelligent systems
- Why the conversation around MCP is as much cultural as it is technical
- How data teams can think about context, confidence, and control as they explore MCP-like architectures
- Where MCP’s potential — and its current limits — might shape the next phase of AI infrastructure
