Merging & Split

Handling support requests in Slack can quickly become overwhelming. Customers often:

• Send messages in rapid bursts

• Change topics mid-thread

• Follow up without clear context

This fragmentation makes it difficult for support teams to track issues, prioritize responses, and maintain continuity in conversations.

To address this, ClearFeed offers two powerful conversation merging capabilities:

• Rule-Based Merging

• AI-Based Merging

Rule-Based Merging

Rule-Based Merging uses a set of deterministic rules to decide when consecutive messages should be grouped into a single request. This ensures related updates stay together, improving context and reducing clutter.

Messages are grouped if they meet one or more of these conditions:

• Same Author: Consecutive messages from the same person

• Short Time Gap: Messages sent within a short window (default: 5 minutes)

• Continuation Cues: Phrases like “but,” “otherwise,” or “on the other hand”

Example A customer might write in 3 different messages on a Slack Request Channel:

“Unable to access your application.”

“We’re getting a 503 error.”

““Any update on this?”

Without merging, these could appear as separate requests. With Rule-Based Merging, they’re automatically grouped, preserving context for faster and more effective responses.

AI thread merging

While Rule-Based Merging handles straightforward cases well, it can miss the subtle ways users continue conversations. ClearFeed’s AI-Based Merging enhances context tracking by applying an intelligent model on top of existing rules.

The AI model identifies when messages are contextually related—even if they're vague, delayed, or indirect.

Key capabilities include:

1. Pronoun & Entity References: Understands vague continuations like

   • “Let me check that.”

   • “Any update on this?

   • “Still investigating it.”

2. Indirect Continuity: Detects when a new participant joins but continues the same topic:

   • “Looking into this now.”

3. Follow-Up & Resolution Cues: Recognizes acknowledgments and contextual replies:

   • “Got it, that solved the issue.”

   • “Here’s the version: v2.1.0”

   • “Uploading the screenshot now.”

4. Urgency & Escalation Detection: Identifies urgency even when the issue isn’t restated:

   • Original: “We’re facing some issues with our prod cluster.”

   • Follow-up: “This is quite urgent, could someone please check!!!”

Important: To prevent reviving old threads unnecessarily, AI-Based Merging applies a 3-day cutoff. Any message sent more than three days after the last activity will start a new request, even if it seems related.

Messages in Request channel

Merge Settings in Your ClearFeed Account

• AI-Based Merging is the default for all new accounts and collections.

• Rule-Based Merging remains active for older collections.

UI configuration is not yet available; all changes are managed by ClearFeed team. If you'd like to switch merging modes, contact the ClearFeed support team.

Splitting Unrelated Messages

Merging helps keep context together, but sometimes, customers report unrelated issues in a single burst. Automatically merging such messages could create confusion.

To solve this, ClearFeed supports Request Splitting.

With Request Splitting, agents can manually separate unrelated messages into distinct requests. This ensures each issue:

• Has its own clear thread

• Can be assigned and resolved independently

• Is tracked without context bleed