Skip to main content

Upload audio

This guide explains how to manually upload one or more audio files in AuditorIA. When you start the upload, AuditorIA creates a processing task so the call can be transcribed and reviewed.

If your team loads audio through an integration, you may not need this screen. In that case, search for the audio from Search audio or find the created task from Search tasks and audio.

audio upload form with campaign, operator, call start, direction, and file upload area

Open Upload file

In the sidebar, click Upload file. The screen opens with the title Upload files.

The section may not appear if your user does not have permission to upload files manually.

Complete the details

Before attaching the audio, complete the main call details.

FieldWhat to enter
Campaign IDCampaign the audio belongs to. It defines the audit context.
Operator IDPerson being evaluated in the call. It becomes available after choosing the campaign.
Call startDate and time when the call started.
DirectionWhether the call was Inbound or Outbound.
info

If you do not see the campaign or operator you need, it may depend on your permissions or the campaign configuration.

If those records do not exist yet, they need to be created or assigned from the corresponding sections first: Campaigns, Local operators, and Quality scorecards.

Attach files

Under Attach files, click the upload area or drag files onto the screen.

You can upload:

  • one audio file;
  • supported audio or video files;
  • a bulk upload with ZIP + CSV.

After you attach files, AuditorIA shows per-file configuration. Check that each audio has the correct campaign, operator, date/time, and direction. If all files share the same values, you can use Apply global values to all.

per-file configuration after attaching several audio files

warning

The current limits are 100 MB per file and 100 MB per batch. If a file exceeds that size, the upload may fail.

Bulk upload with ZIP + CSV

To upload several audio files together, prepare two files:

  • a ZIP with the audio files;
  • a metadata CSV with the information for each audio file.

The CSV does not contain the audio. It tells AuditorIA which campaign, operator, date/time, and direction correspond to each file included in the ZIP.

To use it:

  1. Click Download metadata CSV template.
  2. Complete one row for each audio file you will include in the ZIP.
  3. Keep the column names from the template.
  4. Create the ZIP with the audio files.
  5. Attach both the ZIP and the CSV in the upload.

The template includes these columns:

ColumnWhat to enter
file_nameExact name of the audio file.
operator_idID of the operator or evaluated person.
campaign_idCampaign ID.
inicio_llamadoDate and time when the call started.
directionCall direction, for example ENTRANTE or SALIENTE.
tip

The value in file_name must match the audio file name inside the ZIP. If it does not match, AuditorIA cannot apply that row to the correct file.

Advanced options

The screen includes Advanced options for language, device, and model.

For normal use, leave them as they are unless your team tells you to change something. These options are for specific cases where audio processing needs adjustment.

advanced upload options with language, device, and model

Start the task

When the details are complete:

  1. Review campaign, operator, date, direction, and attached files.
  2. Click Start task.
  3. Wait for the upload confirmation.

If you uploaded one file, you will see a confirmation such as Task sent. If you uploaded several files, the screen may show how many tasks were queued.

What happens next

After you start the task, AuditorIA processes the audio to generate the transcription and the information needed for review.

While processing is not finished, the task may appear as pending or in progress. When it finishes, you can open the task and review the transcription.

If the audio was already loaded

Some audio files may enter AuditorIA through integrations or previous uploads. In those cases, you do not need to upload them again from this screen.

Use Search audio to find audio records by campaign, operator, date, or direction. In that table, you may see actions such as:

  • Transcribe audio, when the audio is loaded but does not have a processed task yet;
  • View task, when the audio already has an associated task and can be opened.

Those actions belong to the audio search flow. This page only mentions them so you know how to continue if the audio already existed.

Continue