Workers
What workers are and how to add or customize them.
- What is a worker
A Worker is an always-on automation trained on your knowledge that starts as a copilot, then can go live on its own.
- Add a custom worker
Create a Support chatbot for your Crisp inboxes, or a Success worker for onboarding, upsell, and churn, with the Add Worker wizard.
- Enable or pause a worker
Turn a Worker on once its knowledge, Skills, and a test run look right, or pause it any time without losing its setup.
- Duplicate a worker
Copy an existing Worker into a new, disabled instance so you can try a second playbook without starting from a blank wizard.
Worker settings
Triggers, actions, and input settings.
- Add an action to a worker
Pick the actions a worker may take — like sending a Slack message or updating Crisp — and fill in the required settings.
- Let a worker browse a public URL
Add the browse-URL action when a worker needs to read a public page that isn't in your Knowledge base.
- Add a trigger to a worker
Add a live or scheduled trigger so a worker only runs on the events you choose.
- Schedule a worker to run on a cadence
Set a repeating schedule and timezone on a trigger, then check the next run time on the worker.
- Limit a trigger with customer segments
Add a segment condition so a trigger only continues for the accounts or messages that match.
- Wake a worker when another worker finishes a result
Add a Worker result trigger so one worker runs each time another finishes a Result, like a VoC worker that analyses every closed chat.
- Add or edit Guards
Add misbehaviour rules on Guards so drafts can't skip the policy you care about.
- Edit Agent Skills for a worker
Open Knowledge → Agent skills to edit the playbook a worker reads at runtime.
- Save and restore skill versions
Save a Skill as a new version before a risky edit, or restore an older version if a change doesn't work out.
- Set worker input fields
Define typed input fields so Zentesimal maps spreadsheet rows, JSON, or free text into them before the worker runs.
- Use Datadog RUM as worker input
Connect Datadog and map RUM unique users into a scheduled cohort for batch workers.
Running workers
Run a worker, read results, and work in batches.
- How to run a worker
Use Run Worker for tests and on-demand runs. Choose Batch or Single event, pick an input, then check Results. Ongoing work uses Triggers.
- Read and filter Results
Open the Results tab, filter by date or column, then read each row for the full outcome.
- Add a note on a Result
Leave an append-only note on any Worker Result so the next teammate who triages it has context.
- Add a prompt column on Results
Add a prompt column when you need a one-off label or summary on existing Results.
- Reprocess Crisp conversations with a Support worker
Rerun your Support worker on Crisp chats still waiting on a reply, in windows of up to 24 hours, from Run Worker → Batch.