Skip to main content

Voice to quote request

Turns a recorded voice memo into a quote request in Propeller. A sales rep records "Mark from ASML wants 10 steel pipes and 4 clamps" after a call. A quote request with the right customer and products appears for review.

Runs: Outside Propeller · Trigger: Webhook · Difficulty: Advanced

What you need​

  • An n8n account. See Set up n8n.
  • An OpenAI account, for the transcription and the matching.
  • A Propeller API key.
  • A recorder that sends the audio file to the webhook with a username and password (basic auth).

How it works​

You record a voice memo and your recorder sends it to the workflow's webhook. The workflow then:

  1. Transcribes the recording and lets AI pick out the customer, the company, remarks and the requested products.
  2. Finds the customer by the names in the memo. The AI picks from at most 20 candidates.
  3. Matches each requested product on its article number first. Otherwise it searches the catalogue and the AI picks from at most 10 candidates per product.
  4. Creates a quote with the remarks, sets it to the status in ORDER_STATUS (default REQUEST, as in the earlier version) and creates a ticket in the Action Hub. Products that could not be matched are listed on the ticket. When two lines match the same product, their quantities are added up and the ticket lists both lines, so you can check the quantity.

Every memo ends in exactly one ticket. When no quote request can be made, the ticket says why and shows the start of the transcript.

Install it​

  1. In n8n, import the downloaded JSON as a new workflow.
  2. On the Webhook node, set a basic auth credential with a username and password you choose, then copy the production URL.
  3. Connect your OpenAI account on the four AI nodes: Transcribe Recording, Extract Request Data, Match Customer and Match Products.
  4. Create a header auth credential with your Propeller API key and select it on the eight GraphQL nodes.
  5. In the Settings node, set CHANNEL_ID, ASSIGN_TO_ADMIN_USER_ID, LANGUAGE (the language of your product names) and TRANSCRIBE_LANGUAGE (the language of the recordings).
  6. Set up your recorder to send the audio in a field called audio to the webhook URL, with the same username and password.
  7. Record a test memo and check the Action Hub.
  8. Activate the workflow.
Filling in your API key

In n8n, create a credential of type Header Auth and fill in its two fields exactly like this:

FieldWhat to fill in
NameapiKey
ValueYour Propeller API key

Name is not a title you pick yourself. It has to say exactly apiKey, otherwise the workflow cannot sign in. See Set up n8n for the full steps.

Make it yours​

  • Languages: LANGUAGE and TRANSCRIBE_LANGUAGE in Settings.
  • Status and channel: ORDER_STATUS and CHANNEL_ID.
  • The AI model: pick another model on the AI nodes.

Limits​

Keep memos short. Long recordings are limited by the request size n8n accepts and the file size the transcription service accepts.

A run that takes longer than one hour is stopped and shows as canceled in the n8n executions list.

Use it outside n8n​

Copy the JSON into an AI assistant and follow Adjust a workflow with an AI assistant.

Good to know

The workflows and agents in the Agent Hub are free to use and adapt. They are examples, not supported product features, so test them in a staging environment and check the results before using them in production.