Front and Knowledge: Policy Refresh from Conversations Template

A Spojit template that keeps a Knowledge collection current by embedding the canonical answers from recently resolved Front conversations on a schedule.

What It Builds

A Schedule trigger fires on a recurring cadence (for example, every night). A Front Connector node pulls conversations resolved since the last run, a Connector node in Agent mode distills each thread down to the question and the final answer your team gave, and a Knowledge node embeds those answers into a persistent collection. Future workflows that query that collection then draw on your latest support resolutions.

The Prompt

Paste this into Miraxa and it builds the workflow, connecting the tools for you:

Build a workflow that runs every night at 2am. Pull all Front conversations that were resolved in the last 24 hours, and for each one summarize the customer question and the canonical answer my team gave. Embed those question-and-answer pairs into a persistent Knowledge collection called "Support Answers" so future replies can draw on the latest resolutions.

Connectors Used

  • Schedule trigger - runs the refresh on a recurring cron cadence in your timezone.
  • Front - reads recently resolved conversations and their messages.
  • Knowledge - embeds the distilled answers into a persistent collection for later querying.

Customize It

Change the cron time and the lookback window (24 hours) to match how often your team resolves tickets. Rename the Support Answers collection, or filter the Front pull to a specific inbox or tag so only customer-facing resolutions get embedded.

Tips

  • Give the Front connection read scope on conversations and messages so it can fetch full threads.
  • Embed into a persistent collection (not Transient) so each run adds to the same searchable body of answers.
  • Use Agent mode for the summarize step so it can pick out the real answer from a long thread.

Related

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.