Restaurant Data MCP: foodservice data for any AI assistant or agent
The Dashmote Restaurant Data MCP plugs licensed restaurant and foodservice data straight into the AI assistants and agents your team already uses. Outlets, menus and prices, beverage brands on menus, delivery app listings, reviews, social signals and lead scores, covering 60+ countries. It works with frontier models such as Claude, GPT, Gemini and Microsoft Copilot, with local and open models, and with agent harnesses such as OpenClaw and Hermes Agent (Nous Research).
The trusted intelligence layer for teams selling to Hospitality & Retail
What an MCP server is, in plain words
The Model Context Protocol (MCP) is an open standard that lets AI assistants and agents connect to outside data. Think of it as a universal plug: once a data source speaks MCP, any assistant or agent that supports MCP can use it, without a custom integration for each one.
For a foodservice team, that means your assistant can answer 'which outlets list our brand on one delivery app but not another?' from licensed restaurant data, instead of searching the web and guessing. The assistant does the reasoning; Dashmote supplies the facts.
Works with the AI your team already uses
Teams do not standardise on one AI tool, and they should not have to. The Dashmote Restaurant Data MCP connects to any MCP-capable assistant or agent harness that accepts a connection key, and apps that connect only through a personal sign-in are available on request:
- Frontier models: Claude (Claude Code, Claude Desktop, Claude for Team and Enterprise), agents built on OpenAI's GPT models, Gemini CLI and agents built with Google's ADK, GitHub Copilot and Copilot Studio.
- Local and open models: run the model on your own machines and keep prompts in-house, while Dashmote supplies the restaurant and delivery data.
- OpenClaw: the free, open-source personal AI agent that runs on your own machine, takes instructions through messaging apps and connects to models such as Claude, GPT, Gemini or local models.
- Hermes Agent (Nous Research): the open-source agent that supports MCP servers and lets you switch model providers.
- Any other agent framework that can connect to an MCP server, including the ones your own engineers build.
Pick the assistant that suits each team. The restaurant data underneath stays the same, so answers agree across tools.
What your agent can ask
Once connected, your assistant can reach every kind of Dashmote data your workspace licenses: outlets, delivery listings and availability, menus and prices, beverage brands on menus, reviews, social signals and lead scores. The questions below are typical of what eCommerce and field sales teams ask on day one.
Answers depend on the countries, platforms and data your workspace licenses.
For eCommerce managers: your delivery shelf in a chat
Ask your assistant the questions you would otherwise send to an analyst and wait a week for:
- 'Which key accounts in London list our brand on one delivery app but not the other?'
- 'How has the price of our hero product moved across platforms since the last refresh?'
- 'Which outlets dropped our listing this month?'
The assistant answers from licensed delivery data, refreshed monthly, and can turn the answer into a list, a summary for the account team, or a follow-up task. Your digital shelf becomes something you check whenever you need to, not once a quarter.
For field sales reps: a copilot that knows the territory
A rep can message an agent such as OpenClaw from their phone, or ask Copilot or Claude at their desk, and get outlet-level answers before the first visit of the day:
- 'Which bars near me do not carry our brand yet?'
- 'What is on the menu at this restaurant, and which drinks do they list?'
- 'Rank the outlets in my patch by fit and tell me why.'
The answers come from Dashmote outlet, menu and lead score data, so the rep walks in prepared and the manager knows coverage is planned on facts.
Governance for agents, in plain words
An agent that can reach a data source will use it often. Dashmote keeps that safe:
- Licensed data only. Your assistant answers from data Dashmote licenses and maintains, not from whatever it finds online.
- Scoped per workspace. Each workspace sees only the countries, platforms and data it licenses. Another company's data is never in your answers.
- Dated and defined. Every answer carries its refresh month and says what it counts. An outlet is a physical venue; a storefront is that venue's listing on one delivery app, so the assistant does not add one to the other.
- Consistent across tools. Claude, GPT, a local model and your own agent all see the same records, so answers agree whichever tool a colleague prefers.
The business case for an AI-native team
Why put Dashmote in your AI stack:
- Head of eCommerce: delivery listings, prices and availability your team can check in a conversation, by platform and city.
- Head of Sales: reps with a copilot that ranks outlets and briefs them before each visit.
- Data or AI lead: one restaurant data source for every assistant and agent harness, frontier or local, with no custom integration per tool and no vendor lock-in.
- Everyone: answers grounded in licensed, dated data instead of web search guesses.
API or MCP: same data, your choice
The MCP route is the fastest way to put Dashmote inside an assistant or agent harness. If you also want the data in your own apps, sales tools or warehouse, the Restaurant Data API gives you the same licensed data. Many teams use both: MCP for their assistants, the API for their products. For the full list of supported AI tools, see AI connectors.
How to get access
There is no public sign-up. Tell us which assistant or agent harness your team uses and which questions it must answer. Our team sets up the connection for your workspace, scoped to the countries, platforms and data you license, and checks that your first questions come back right.
Frequently asked questions
Which AI tools does it work with?
Any assistant or agent harness that supports MCP with a connection key, and our API for everything else. That includes Claude Code, Claude Desktop and Claude for Team and Enterprise, agents built on OpenAI's GPT models, Gemini CLI, GitHub Copilot and Copilot Studio, agent frameworks such as OpenClaw and Hermes Agent (Nous Research), and agents your own team builds. Apps that connect only through a personal sign-in, such as the ChatGPT app, are on request.
Does it work with local models?
Yes. If you run an open or local model through an MCP-capable harness, it can use Dashmote data the same way a frontier model does. The model runs where you choose.
Do I need developers?
No. Our team sets up the connection for your workspace and your assistant. Your team then asks questions in plain language.
Is the data licensed?
Yes. Dashmote data is licensed per workspace, and your assistant answers only within what your workspace licenses. Every answer carries its refresh month and definition.
Is it the same data as the Dashmote API?
Yes. The MCP route and the Restaurant Data API give you the same licensed data under the same workspace licence.
How do I get access?
Request access and tell us which assistant or agent you use. There is no public sign-up; our team sets up the connection for your workspace.