Skip to content

Import recordings

The Import call recordings job ingests audio files and metadata produced outside the platform — a third-party recorder's archive, a MiaRec export from another system, or a directory of bare audio files. Imported recordings become regular conversations and enter the same processing pipeline as recorded ones.

Where: Administration › Storage › Import Recordings, which opens the jobs console filtered to this job type. Create a job, configure it, and run it manually or on a schedule.

Job settings

  • Data source — how the job tracks progress across runs: with a continuation token (each run picks up where the previous one stopped) or full mode (re-scan everything).
  • Vendor Format — the metadata format of the source archive: MiaRec (*.json) (the format written by export jobs and replication), SmartRecord (*.meta), Verba (*.xml), Avaya ACR (*.xml), Crystal Reports (*.xml), RedBox (*.txt), Switchvox (*.xml), or bare MP3 files (*.mp3) / WAV files (*.wav). For bare audio files, a REGEX pattern extracts metadata (dates, phone numbers) from the file names.
  • Source Storage, Source Directory, Source File Mask — where to read from; the source is a storage target.
  • Destination Storage, Destination Directory, Filename format — where the imported audio files are written and how they are named (the parameterized format described in Files location).
  • Tenant (optional) — assign all imported recordings to one tenant. Without it, the importer matches tenants from the metadata; enable Ignore unknown tenants ("Do not import recordings for unknown tenants") to skip records it cannot match.
  • Metadata only — import the call records without copying audio files.
  • Remove source files — delete source files after a successful import (not available together with metadata-only).

The randomization settings (random IDs, phone numbers, dates) exist for generating demo/test data sets from an archive — leave them disabled for production imports.

Run progress, per-record results, and logs are available on the job page like for any job — see Jobs console.

Files on network shares

Source files are read by the MiaRec server processes, not by your browser — a path that opens fine from your workstation can be unreachable from the server. If the archive sits on a network share, either mount the share into the server's local file system (for example, /mount/backup) or register it as an SMB storage target, and make sure the service account running the MiaRec background workers has access to it.

Importing encrypted recordings

A MiaRec-format archive exported from an encrypted system contains encrypted audio. Import the matching encryption key before (or along with) the import, or the recordings cannot be played — see Encryption.