Sharing a Workspace Without Sharing Data: fastn UCL's Smart Multitenancy Setup

Multitenancy is the ability for a single platform instance to serve multiple customers or user groups while keeping their data, tools and configurations completely separate. It is essential for building scalable, secure systems where each tenant operates independently without needing duplicate infrastructure.

What is multitenancy in the Unified Context Layer?

At its core, multitenancy means a single deployment of the UCL platform can serve multiple tenants, such as different customers, business units or user groups, while keeping their data, integrations and permissions completely separate.

Architecture overview

UCL's multitenant architecture is powered by a centralised MCP-compatible server that dynamically routes every action based on tenant context.

• •Each tenant's tokens, actions and configurations are stored separately.
• •A shared connector infrastructure ensures reuse and modularity.
• •Context-aware execution logic ensures that actions are routed only to the tools and permissions associated with the calling tenant.
• •The platform supports horizontal scaling, so you can onboard new tenants without worrying about performance impact.

This is achieved without duplicating infrastructure, making UCL both cost-efficient and easy to scale.

How multitenancy works: a simple example

In many real-world apps, multiple users, such as teammates from the same department, collaborate in a shared workspace. For example, Sarah, June and Emily might all belong to the marketing team and operate under the same Space ID. While they appear to be working in the same environment, UCL ensures their actions, configurations and data remain securely separated behind the scenes.

UCL assigns each user a distinct tenant context. This means when Sarah connects to Slack or Emily accesses Notion, they are only interacting with their own tools and credentials. No extra logic is needed on your part.

Tenant context keeps everything organised

When Sarah logs in, UCL recognises her tenant identity. She sees only her tools, such as Slack, HubSpot and Salesforce, and any action she triggers, such as sending a message, uses only her configurations. June and Emily have the same experience, completely independent from each other, even though they are in the same shared workspace.

Same connectors, different usage per tenant

All tenants in the workspace have access to the same set of available connectors, but they configure them independently.

• •Sarah connects Slack and enables the Post Message action.
• •June connects Slack as well, but disables messaging and only uses status updates.
• •Emily connects Notion and enables only read-only access to the documentation.

Each tenant's integrations, tokens and actions remain completely isolated.

UCL handles the complexity for you

You do not need to write custom tenant logic or routing. fastn UCL automatically:

• •Reads the tenant ID from every command
• •Matches the request to the correct connector configuration
• •Executes only permitted actions based on role and workspace level access
• •Logs and monitors activity per tenant

This lets your team focus on building features while UCL ensures secure and reliable multitenancy behind the scenes. Monitoring in UCL provides a real-time view into how requests flow through your integration layer, helping you track performance and tenant-specific operations.

Why this matters

• •Security: complete tenant isolation prevents data leakage or unauthorised access.
• •Simplicity: users manage their own tools and tokens without requiring system-wide changes.
• •Scalability: one deployment supports thousands of tenants without performance tradeoffs.
• •Efficiency: shared infrastructure reduces costs while maintaining privacy.
• •Central management and ownership: admins can oversee all tenants while respecting boundaries.

If you are building an AI native application that interacts with external tools on behalf of your users, multitenancy is essential. UCL gives you a production-ready, secure foundation to serve multiple users, organisations or workspaces, all from a single integration layer.

No manual routing. No risk of data exposure. Focus on features, and let fastn take care of the complexity.