DoorDash has introduced a connector that allows companies to integrate its delivery service into internal AI tools, enabling employees to place and track food and supply orders through workplace platforms.

The connector uses the Model Context Protocol (MCP), an open standard for connecting AI applications to external services. It gives compatible AI agents the ability to search DoorDash, build a cart, place an order, and track delivery. Companies can use it with existing tools or build their own agents.

SpaceXAI, Vercel, Cognition, Tempo, and Mercor are among the early beta partners using the connector for employee ordering, according to a press release from DoorDash.

Potential use cases described by the company include a Slack bot that collects lunch requests from a team thread and places a group order, as well as an inventory-connected agent that automatically reorders office snacks or supplies through DoorDash when stock runs low.

Access is managed at the company level, with individual employees required to sign in to their own DoorDash accounts and agree to terms before an agent can act on their behalf.

The current phase is a limited beta focused on corporate employee ordering. Beginning September 30, companies can join a waitlist for broader beta access. DoorDash said it is also expanding beta access to its command-line interface for individual developers building agents for personal use.