What is Model Context Protocol (MCP) in corporate travel?

Model Context Protocol (MCP) is a connectivity standard that allows AI agents to plug into external systems - such as booking tools, policy engines, and traveler profile databases - through a common interface, without requiring a custom-built integration for each system. In corporate travel, MCP enables AI agents to discover and interact with travel data sources in a structured, consistent way. Anthropic introduced MCP in 2024, and adoption across the travel technology sector is accelerating.

How does MCP affect TMCs?

MCP strengthens the TMC's position in an agentic world rather than threatening it. Because a TMC holds traveler profiles, policy rules, preferred rates, duty of care data, and payment infrastructure in one place, its MCP layer can deliver a far richer experience than any single supplier endpoint. TMCs that build MCP infrastructure around this intelligence - exposing a genuine policy engine alongside negotiated content - become the layer that AI agents rely on.

Does MCP make NDC obsolete?

No. NDC and MCP solve different problems and will coexist in most travel technology architectures. NDC addresses a content distribution challenge: enabling airlines to share richer offers through intermediaries. MCP addresses a connectivity challenge: giving AI agents a standard interface to reach those systems. If anything, MCP should make NDC content more accessible by pushing airlines and GDSs to simplify how they expose their APIs.

What does MCP mean for the GDS model?

MCP introduces genuine structural pressure on the traditional GDS transaction-fee model. If airlines and hotels expose their own MCP servers directly, AI agents can query them without routing through a GDS. TMCs using MCP connections could simultaneously query a GDS, an OBT, and a direct airline connection — making pricing transparency across every channel unavoidable. GDSs that reposition as trusted aggregation layers have a strong hand to play; those that respond with locked-in, proprietary setups risk fragmenting the standard.

What should TMCs do to prepare for MCP?

TMCs should take two immediate steps. First, audit data quality — not just where data lives, but whether traveler profiles, policy documentation, and preferred rate structures are clean enough for an AI agent to act on reliably. Second, get policy into machine-readable formats. Intent documented in a Word file works for a human agent who can interpret it; an LLM needs structured, deterministic rules to enforce policy consistently. Companies that complete this work early will extract significantly more value when MCP tooling arrives at scale.

How does Oversee relate to MCP and agentic AI?

Oversee operates as an AI optimization layer embedded in the workflows TMCs already use. Across air and hotel, the platform continuously monitors booked travel, identifies savings opportunities, and acts on them automatically — without rebuilding existing operations. AgentSee extends that logic to agent productivity, automating ticket handling and service workflows. As MCP becomes the standard interface between AI tools and travel data, having clean, well-integrated optimization layers already in place compounds in value. TMCs already running Oversee are better positioned to benefit from agentic infrastructure as it matures.

What is an AI optimization layer for travel?

An AI optimization layer is a platform that sits across existing travel operations and continuously improves outcomes - capturing savings, automating workflows, and surfacing intelligence - without replacing the underlying systems TMCs already use. Rather than requiring a workflow rebuild, it plugs into the existing tech stack and operates in the background. Oversee is an example: connecting to GDS and booking environments to monitor and act on air and hotel bookings at scale.

How does agentic AI differ from standard travel automation?

Standard automation executes predefined, rule-based tasks. Agentic AI can discover what a system can do, reason about options, and take multi-step actions on behalf of a user - adapting based on context rather than following a fixed script. In corporate travel, this means an AI agent working through MCP can query traveler profiles, check policy, identify preferred rates, and initiate rebooking - all within a single workflow, without a human initiating each step.

Oversee blog