IMPORTANT: This feature is available in public preview mode.
What is MCP?
Model Context Protocol (MCP) is an open standard developed by Anthropic that defines how AI applications communicate with external tools and data sources. It works like a bridge: any MCP-compatible AI client can connect to an MCP server, which in turn provides access to tools, data, and documentation from a third-party system.
What the Synerise MCP Server enables
The Synerise MCP Server is an implementation of the MCP standard for the Synerise platform. Once configured, you can connect any MCP-compatible AI client to your Synerise workspace (a chat assistant like Claude, a coding agent like Claude Code or Codex, or your own custom agent) and query analytics, inspect campaigns and workflows, retrieve customer profiles, explore AI models, browse loyalty programs, and access developer documentation, all in the flow of your work. If write access is enabled, you can also create and update resources, such as analyses, dashboards, documents, and AI recommendation campaigns.
This removes the need to:
- Export data manually and paste it into your AI client
- Navigate through multiple Synerise screens to retrieve figures
- Search for documentation separately while working in your AI client
You get what you need from your workspace in real time, in the context of your request.
What is available
The Synerise MCP Server covers the following product areas:
| Product area | What you can do |
|---|---|
| Analytics | Metrics, segmentations, funnels, trends, histograms, Sankey diagrams, reports, aggregates, and expressions, including live calculation and profile segmentation membership checks. Creating and updating saved analyses, and managing dashboards, their widgets, sharing, and folders |
| Campaigns | Campaign details, templates, in-app messages, landing pages, documents, and screen views from the Experience Hub. Creating and configuring documents and screen views |
| Automation | Workflow details, schedules, statistics, execution history, upcoming triggers and status changes, workflows with issues, in-progress workflows, performance rankings, workflow notes, status counts, data transformations, and workflow and node template folders |
| AI Recommendations | AI recommendation campaigns, AI Engine configurations, statistics, top product performance, and live recommendation fetching and previewing. Creating, updating, copying, and changing the state of AI recommendation campaigns |
| AI Search | AI search indices, query rules, synonyms, query statistics, top queries, recent searches, live search query execution, autocomplete, and filter-only browsing. Creating and configuring AI search indices, suggestion indices, query rules, and synonyms |
| AI (other) | AI predictions and AB tests, including creating, updating, and copying them, and calculating the AB test sample size. Finding AI recommendation campaigns and AI search indices by name |
| CRM | Customer profile lookup and contact search |
| Loyalty | Promotions, voucher pools, and personalized promotions (handbills), including creating and updating them and importing promotions in bulk |
| Assets | Catalogs, Brickworks schemas, records, external sources, and personalizations, code snippets, and API keys. Generating and previewing Brickworks records, creating catalogs and adding items, and configuring catalog filtering |
| Events | Event definitions and parameters, and the event history of a profile. Registering events, adding parameters, and managing their visibility and parameter defaults |
| Account | ID and name of the connected workspace |
| Mobile SDK guidance | iOS and Android SDK setup guides, configuration help, best practices, diagnostics, FAQ, integration validation |
| VTEX guidance | VTEX plugin setup, component guides, configuration help, integration validation |
For a full list of available operations, see Available tools.
Example workflows
The following examples show how different team members can use the Synerise MCP Server to get more done in less time.
Analyst: querying data without switching tools
An analyst wants to understand drop-off in the checkout funnel over the last 30 days. Instead of opening Synerise, building the funnel query, waiting for results, and then copying figures into a presentation, they ask directly:
Calculate the checkout funnel for the last 30 days. Show me the steps and where users drop off most.
They get back a structured breakdown, including the step with the highest drop-off rate and a suggested hypothesis.
Analyst: building a dashboard
The same analyst wants to track the funnel together with revenue and share the results with the marketing team. With the Write action enabled in the workspace, they ask:
Create a dashboard with the checkout funnel and the weekly revenue trend, and share it with the marketing team.
The AI client creates the dashboard with both analyses as widgets and shares it with the selected users, without opening the platform.
Marketer: reviewing campaign status
A marketer preparing for a weekly review wants a summary of all active email campaigns:
List all active email campaigns and tell me which ones are scheduled to send in the next 7 days.
They get the matching campaigns back with their scheduled send times, without opening the platform.
Developer: setting up a mobile integration
A developer integrating Synerise into a new iOS app asks:
Walk me through setting up push notifications with the Synerise iOS SDK.
They get a step-by-step walkthrough, built from the relevant Mobile SDK setup guide, configuration documentation, and best practices, tailored to their question.
Operations: checking automation health
An operations manager wants to know if any workflows have issues before the weekend:
Are there any workflows currently failing or paused unexpectedly?
They get automation status back immediately, including which workflows have issues and what kind of problem each one has, so the team can investigate before going offline.
Who benefits
The Synerise MCP Server is useful for anyone who regularly works with Synerise data or needs to reference Synerise documentation:
- Analysts and data teams - run analytics queries in natural language, explore segmentations, build dashboards, and get results explained in context.
- Marketers - check campaign status, loyalty promotions, and recommendation performance without switching tools.
- Developers - get guided through SDK and VTEX integration steps, validate configurations, and troubleshoot issues.
- Operations and automation teams - monitor workflow health, check upcoming triggers, and investigate execution history.
Data privacy and third party AI clients
NOTE: For answers to questions about the EU AI Act, data processing agreements, and data processing locations, see Synerise MCP Server FAQ.
The Synerise MCP Server operates within Synerise's infrastructure and is covered by your organization’s existing agreement with Synerise (including the applicable data processing agreement). Synerise provides the MCP Server as an integration interface and does not determine whether, how, or for what purpose you connect the Synerise MCP Server to any third-party AI client or AI agent, including Claude, Cursor, Windsurf, or VS Code with Copilot.
If you choose to connect the Synerise MCP Server to a third-party AI client/AI agent, any data the server retrieves - including personal data such as customer profiles, contact details, or event history – may be transmitted to and processed by that AI provider in order to generate a response. Such processing is performed outside Synerise's infrastructure and outside Synerise's control. All terms applicable to that processing are determined by the provider of the relevant AI client or AI agent and by your configuration and use of that AI client/AI agent.
By connecting any third-party AI client or AI agent to the Synerise MCP Server, you remain solely responsible for assessing, authorizing, configuring, and monitoring that connection. Synerise does not verify or approve third-party AI clients/AI agents connected by you and does not control the prompts, instructions, outputs, or downstream processing performed by those tools.
Before connecting the Synerise MCP Server to any third-party AI client or AI agent, confirm that:
- your organization permits the use of it and has identified a valid legal basis for that use, including, where required, an appropriate data processing agreement and other contractual arrangements with the relevant AI provider.
- the intended use case complies with applicable data protection, confidentiality, sector-specific, and AI-related legal requirements, including any restrictions on international data transfers.
- the Synerise account used to authenticate has only the permissions necessary for the intended use case, to limit the personal data and other confidential information that can be accessed and transmitted to the connected AI client or AI agent.
Limiting the allowed actions to Read limits what the Synerise MCP Server can change in your workspace; it does not limit the scope of data, including personal data, that can be read, retrieved, included in prompts or context, and transmitted to the connected AI client or AI agent.
Synerise is not responsible for processing performed by third-party AI clients or AI agents selected, connected, configured, or used by you, including any use of data by those providers, any retention or deletion practices, any model-training or model-improvement use, any generated outputs, or any decisions or actions taken on the basis of those outputs.
Limitations
- Workspace settings - access for external AI clients is disabled by default. A user with the PROFILE_ADMIN role must enable it in the Synerise MCP Server settings, where they also decide whether the server can only read data or also create and edit resources. There are no tools for deleting resources, but some write tools replace or clear existing data, for example, updates that clear the fields you don't send again. For details, see Available tools.
- User permissions - your AI client acts on your behalf, so it can only perform actions that your user permissions allow. Enabling access in the workspace settings is the first step. Users with custom roles also need the AI Agents > MCP Server permission, which an admin grants in the role permission matrix. Basic roles and predefined Synerise roles have this permission by default. For details, see Allowed actions and user permissions.
- Workspace scope - each server connection is linked to a single workspace, which you select when you log in. To work with several workspaces, add the server several times. For details, see Connecting several workspaces.
- No real-time event streaming - the server retrieves data on demand in response to each question; it does not push live updates or subscribe to workspace events.
- AI client behavior - the accuracy and usefulness of responses depends on how your AI client interprets your request and selects tools. Complex or ambiguous requests may require follow-up clarification.
What's next
Proceed to Getting started to configure the Synerise MCP Server with your AI client.