Connect your tools

How integrations feed the three senses

Unless connects to the systems your business already runs, and every one of those connections feeds one of three senses: Living Knowledge, Living Memory, and Living Context. These three senses are what let the agent know what it knows, who it is talking to, and what it can actually do. Without them, an agent is just a chatbot with no context. With them, every interaction starts from the full picture instead of a blank slate.

The three senses explained

The three senses work together as one mind. Living Knowledge is what the agent knows, built from your documentation and content. Living Memory is who the agent is talking to, drawing on CRM data and past interactions. Living Context is where the agent is acting, connecting to your business systems so it can take real actions.

Living Knowledge: what the agent knows

Living Knowledge pulls from 100+ integrations, including tickets, support conversations, internal documentation, wikis, Slack, Confluence, Google Drive, support recordings, and website content. All of this is continuously restructured and rewritten into one non-ambiguous, always-current library. The same library feeds three surfaces: the Customer Agent, the Team Assistant, and the Help Center. Documentation operations drop to near-zero because the library maintains itself from accepted drafts, overrides, and edits.

Living Memory: who the agent is talking to

Living Memory unites structured data and working memory. It pulls from 40+ CRM and customer data integrations to build a per-customer profile that includes preferences, history, sentiment, and goals. This data is private per customer and scoped per agent. When a customer reaches out, the agent recalls previous interactions and preferences, so the conversation feels like continuing a relationship rather than starting from scratch.

Living Context: what the agent can do

Living Context connects to 20+ business systems, including CRM, billing, ERP, support tools, and custom APIs. This lets the agent understand the customer’s current situation, access relevant records, and take actions like resetting a password, checking an order, triggering a renewal, or firing an in-app notification. Every procedure carries explicit permissions: what the agent can do alone, what needs a human, and what is off limits.

How integrations work in practice

When a customer reaches out, the agent uses all three senses at once. It finds accurate answers from your documentation through Living Knowledge, recalls previous interactions through Living Memory, and sees the customer’s current account status through Living Context. This combination creates personalized, contextual conversations.

Most customers start read-only, letting the agent answer questions and summarize accounts without changing anything. Write actions, like filing tickets, updating profiles, and changing plans, are turned on later, action by action, with guardrails your risk team is comfortable with. You decide which fields the agent can read, which it can write, and which it cannot see at all.

Setting up and managing integrations

Most integrations are pre-built connectors. IT reviews permissions, configures authentication, and decides which data and actions Unless can access. The configuration is small and most of it is reversible from the dashboard. When an integration fails to sync, you will see an error notification in the Unless dashboard, making it easy to spot connectivity issues without checking each integration manually.

If an integrated system goes down, the agent degrades safely. It uses whatever it can still reach, tells the customer if a system is unavailable, and either offers to follow up or hands off to a human. You can configure the fallback behavior per integration.

Security across integrations

Integrated systems inherit the same security as the core platform: encryption in transit and at rest, role-based access, EU residency, and the Privacy Vault for sensitive identifiers. You can keep test and production separate, route data through EU-only paths where regulation requires it, and scope integrations per business unit or region. Different agents can carry different permissions on the same underlying system.

Custom integrations and MCP servers

Unless supports custom integrations through its APIs and connector framework, letting your team or a partner add new systems. Unless Studio is available for hourly work if you want the Unless team to build it with you. Unless also supports MCP servers, which let you expose internal tools, custom APIs, or proprietary data sources to the agent through a standard protocol. The agent reads and acts through MCP the same way it does through any other connector, with the same audit trail.

Conclusion

Integrations are the backbone of the three senses. They turn a generic chatbot into an agent that knows your business, remembers your customers, and acts inside your systems. Whether you start with a few read-only connections or build a full ecosystem, every integration you add makes the agent more useful and more contextual. The result is one agent, one memory, every channel - and a customer who never has to introduce themselves twice.