Unlocking more context for AI: Introducing Merge’s Chat Unified API

Institutional knowledge doesn’t just live in docs. It lives in conversations across Teams, Slack, and other chat platforms.

AI agents and enterprise search tools need access to that context to be useful. But chat data is fragmented, hard to query, and nearly impossible to integrate at scale with the right freshness and permissions.

Today, that changes. We’re introducing Merge’s Chat Unified API: a normalized interface across chat platforms that gives modern AI applications real-time access to conversational knowledge with enterprise-grade access controls.

Here’s more on how Chat Unified API works.

One API, five core objects

Building direct integrations with chat platforms is deceptively complex. Each system has its own rate limits, ways of handling conversation threading, and permissioning, differences that compound quickly as you scale.

Our Chat Unified API eliminates these technical challenges. It abstracts away the underlying complexity and provides you with five core object types designed for efficiency and scale:

• Messages: This includes individual posts including root messages and threaded replies

• Conversations: These are the channels where work happens, including public, private, and shared configurations

• Users: This includes internal employees, external collaborators, and bot accounts

• Groups: These are the teams or organizational units that own conversations and define base access

• Members: This is the critical ACL layer that determines who can access which conversations

Security that scales

In an enterprise environment, “who can see what” is critical to corporate governance and security. 

Our access control list (ACL) model is designed for efficiency, particularly in large public channels that might have thousands of members.

• Public channels: Membership is represented by a single Member object (representing the Group) to avoid per-user ACL fanout
• Private channels: We maintain explicit membership lists for granular control
• Auth flexibility: We offer an admin flow for org-wide visibility
• User flow: This is best suited for apps that only need to see what the authenticated person sees

What you can build today

With Merge’s new Chat Unified API, you are enabling access to your customers’ institutional memory.

• Enterprise search: Build tools that surface relevant conversations across your customers’ communication history

• Context-aware AI agents: Give your assistants the ability to understand your customers’ organizational decisions

• Customer support: Connect your customer success application to your customers’ chat platform to monitor customer issues and provide updates

• Near-real-time insights: Your applications stay current with a ten-minute sync cadence

The roadmap: Microsoft Teams, Slack, and beyond

The Merge Chat Unified API initially supports Microsoft Teams integration, but this is just the beginning. Slack is already on our roadmap and we’ll prioritize others based on customer demand. 

We’re also looking ahead to our next major release, where we plan to introduce webhooks for real-time updates and an attachment common model for file syncing.

Important conversations are no longer scattered in inaccessible channels, DMs, or old archives. Now they can be turned into context that powers knowledge at scale, securely for you/your customers.

Interested in joining our Chat Unified API Closed Beta? Contact our team to request more information.