Tagging and Triaging

Let's say we want to tag every conversation in our community Slack workspace with their relevant group label:

• support — Technical support questions and troubleshooting
• sales — Questions about products, pricing, or purchasing
• billing — Billing, invoicing, or payment-related inquiries
• feedback — Product feedback, feature requests, or complaints

We also want to triage the priority of each feedback conversation with: "low", "medium", "high".

from runllm import Agent, Event, SlackListener
from runllm.decorators import entrypoint

@entrypoint(
    listeners=[SlackListener(team_id="T123456")]
)
def tag_and_triage_wf(agent: Agent, event: Event):
    convo = event.conversation

    # This automatically tags the conversation with any of the tags that fit.
    # As more user messages come in, more tags can be applied.
    applied_tags = agent.tag(
        convo,
        options={
            "support": "Technical support questions and troubleshooting",
            "sales": "Questions about products, pricing, or purchasing",
            "billing": "Billing, invoicing, or payment-related inquiries",
            "feedback": "Product feedback, feature requests, or complaints"
        }
    )
    print(f"Tags applied: {applied_tags}")

    # All tags are automatically saved on convo.tags.
    if convo.tags:
        agent.categorize(
            convo,
            categories={
                "low": "Minor suggestions or general comments that don't require follow-up",
                "medium": "Feedback that could improve product quality or user experience and may warrant review",
                "high": "Critical bugs, urgent complaints, or issues affecting customer retention"
            },
            default="low",

            # This automatically pins one of the categories to the conversation
            # as a tag. This means a conversation can never accidentally get tagged
            # with multiple prioritizes, even as we reevaluate the category.
            apply_as_tags=True
        )

tip

After these tags are applied, you will be able to view, search, filter, and aggregate them in the RunLLM dashboard.