Automate Post-Call Workflows with the Transcript Generated Trigger

By Sparkforge Support Team ·Last reviewed May 3, 2026

The Transcript Generated trigger fires a Sparkforge workflow as soon as a call transcript is ready — for Voice AI, IVR, or Sparkforge Phone calls — and passes the full transcript plus call metadata into your automation.

The Transcript Generated trigger starts a workflow the moment Sparkforge finishes transcribing a recorded call. It works across Voice AI, IVR, and Sparkforge Phone, and passes the transcript, summary, duration, direction, caller location, and timestamps into the workflow so you can automate follow-up the second a call ends.

Where it works

Voice AI calls — transcription is on by default.
Sparkforge Phone calls — transcription must be enabled in Phone Settings.
IVR calls — transcription must be enabled in IVR Settings.

If Voice AI Translation is enabled on the agent, the workflow also receives a translated transcript and translated summary, so you can route localized follow-ups or write standardized outcomes to the CRM.

Why use it

Instant follow-up: trigger SMS, email, tasks, or webhooks seconds after a call ends.
AI-driven branching: feed the transcript into AI Decision Maker or GPT Actions to route on sentiment, intent, or keywords.
Insight capture: write objections, buying intent, or call outcomes back to the contact record automatically.
Rich metadata: branch on duration, direction, timestamps, location, and assigned user.

How to set it up

Open Automation → Workflows and click + Create Workflow (or open an existing one).
Click Add New Trigger and pick Transcript Generated.
Confirm transcription is enabled. Voice AI is on by default. For Sparkforge Phone or IVR, click the CTA in the trigger config to enable it in the relevant settings.
Apply filters: Call Type (Voice AI, IVR, or Normal), Call Direction (inbound/outbound, optional), and Call Duration (minimum seconds, optional — useful to ignore short or abandoned calls).
Click Save Trigger, add the workflow actions, then publish.

Common patterns

Send the transcript to an AI Decision Maker step that classifies the call as booked, missed, or complaint, then routes accordingly.
Update contact custom fields with the call summary so reps see context on the next touch.
Post the transcript to a webhook for revenue intelligence or QA tooling.
Filter on duration > 30 seconds to skip dropped calls.

Frequently Asked Questions

What call types does the Transcript Generated trigger support?

Voice AI, IVR, and standard Sparkforge Phone calls. Voice AI has transcription on by default; the others require it to be enabled in Phone Settings or IVR Settings.

Can I get a translated transcript?

Yes — if Voice AI Translation is enabled on the agent, the workflow receives both the translated transcript and the translated summary alongside the originals.

How do I avoid firing on short or dropped calls?

Use the Call Duration filter inside the trigger configuration to require a minimum length, for example greater than 30 seconds.

What data does the trigger pass to the workflow?

Full transcript, call summary, duration, direction, timestamps, caller location, and the assigned user — usable in any downstream workflow action.

Can the transcript drive AI branching?

Yes. Feed the transcript into AI Decision Maker or a GPT Action step and branch on intent, sentiment, or specific phrases.