Selected work

Workflow automation

Request to Action

Cedar reports a billing issue by email, then follows up in chat. Keep one ticket and one response deadline. If the queue times out, check whether it already saved the ticket before writing again.

Run the intake workflow

Start with the new receipt for BL-184. Then try a timeout after the queue saves a ticket.

Synthetic data

Request details


Routing decision

Loading sample request…

–Queue
–Response target
–Ticket ID

Execution trace

    One ticket keeps its original clock

    Cedar first contacted Billing at 10:00 UTC. A repeat at 14:00 links to BL-184. A follow-up with a second receipt updates BL-184. In both cases Billing keeps the original deadline of 10:00 UTC the next day.

    Download n8n workflow Setup and data notes

    When a write fails

    Check before retrying

    Look up the account and request reference. If a record exists, recover its ticket ID. If no record exists, retry with the original intake time. The trace shows whether a new record was written.

    What a production integration would need

    The demo uses an in-memory queue and elapsed hours. In production, enforce a unique account and reference in the ticket store or use the queue’s idempotency key. Replace the clock with your SLA calendar and define how to review ambiguous or urgent follow-ups.