Set a default language for conversations whose language could not be detected automatically.

Decide how long recognized audio files are kept before automatic deletion.

Enable and tune the disputes workflow.

Control what agents see on dashboards, in tags, and in automated scoring.

Adjust scorecard scoring defaults.

Configure the SMTP server used for invites, password resets, and email reports.

General Settings — the main block of platform-wide toggles and values.

SMTP Settings — the outgoing-email server used for system messages.

Saving any field requires the system-settings management permission. Users who can open the page through the analytics-settings permission alone cannot save: Save changes is rejected by the backend. They also do not see the SMTP card values or other system-only fields — those fields are filtered out of the response and the SMTP card therefore shows blanks or defaults rather than the configured values.

The Default language field requires the Ender Turing staff permission. In cloud (multi-tenant) deployments it additionally requires the workspace superuser flag — a staff user who is not also a workspace superuser still sees the field disabled in cloud. On-premises system-settings managers without the staff permission can see the current value but cannot save a change to it.

The Max days to store recognized audio files field is shown and editable only for users with the system-settings management permission. Other users do not see the field at all.

In on-premises (single-tenant) deployments, every field on the page except Default language is editable for system-settings managers. Default language still requires the Ender Turing staff permission.

In cloud (multi-tenant) deployments, a subset of fields is restricted to a workspace superuser:

In cloud (multi-tenant) deployments, the following fields remain editable for any system-settings manager (no superuser required): First day of week, Default phone prefix, Convert spoken numbers in transcripts, Agent prefix, Agent hourly rate (and its currency), Dispute limit per agent and its period selector (when disputes are already enabled), and Max days to store recognized audio files.

Company logo — pick an image stored in EnderDrive (PNG, JPEG, or JPG) to use as the workspace logo. Click Set as company logo to open the file picker. A preview of the selected file appears below the selector.

Default language — the language Ender Turing falls back to when an audio file does not specify a language and automatic language detection cannot decide one.

First day of week — choose Monday or Sunday. This shapes how date pickers across the platform present weeks (the calendar grid that opens for a date or date-range filter). It does not currently change how weekly report data is grouped.

Default phone prefix — a dial prefix used to broaden phone-number matching and customer lookup. When a search or lookup runs against a local number that has no country code, Ender Turing also tries the prefixed variants so the same contact can be found. The setting does not rewrite numbers as they are stored from incoming calls — the original number is kept; the prefix only widens which stored numbers match a query.

Convert spoken numbers in transcripts — controls whether spoken numbers are converted to digits for transcripts produced through the multichannel diarization path. This covers calls that arrive with separate agent and client channels and are split per channel; it does not currently rewrite transcripts produced through the single-channel fallback path, which stores the speech-to-text result unchanged.

Agent prefix — a string used as a prefix for agent identifiers in your workspace.

Max days to store recognized audio files — how long audio files are kept after recognition before they are automatically deleted. Enter a whole number of days greater than zero; non-integer values are rejected. Conversation transcripts, review scores, feedback, and playlist audio are kept independently of this setting and are not affected by it.

Agent hourly rate — Average Org-wide (Default) — a default hourly cost used in cost-aware analytics when a per-agent rate is not configured. Includes a currency selector (USD, EUR, UAH, PLN, GBP, CAD, CHF, CNY, JPY, KRW, MXN, NZD, SEK, SGD, TRY, ZAR).

Authentication: auto-create Ender Turing users on first OIDC login — shown only when OIDC is configured and the user has system-settings management permission. When enabled, federated identity provider users (for example, Entra ID) receive an Ender Turing account on their first successful sign-in. Existing accounts are linked by verified email, and new users receive the Agent role. Save this value with the page's Save Changes action.

Enable disputes — let agents dispute review results. When enabled, two additional controls appear:

Agents can see only own team tags in conversations — restrict the tags an agent sees in transcripts to those of their team.

Allow agents to see automated scoring and feedback — when enabled, agents can see automated scoring results and the feedback attached to them.

Allow selecting multiple teams on dashboards — let users pick multiple teams at once on dashboards.

Enable percentile filter on dashboard for agent scoring — show a percentile-based filter alongside the standard agent-scoring filter on dashboards.

Allow reviewers to choose a scorecard during conversation review — let reviewers choose which scorecard to apply while reviewing a conversation.

Assign default scorecards based on conversation filters — assign the default scorecard automatically from conversation filters.

Maximum allowed score per scorecard point — the maximum value that a single scorecard point can take. Used when validating scoring forms.

Enable preview features — turn on features that are still in preview.

Enable waveform generation — when enabled, the audio-conversion worker generates a waveform file for newly processed audio. The audio player then renders the saved waveform for those conversations. The setting controls generation going forward only: it does not backfill waveforms for audio that was already processed, and it does not hide existing waveforms — conversations whose waveform file was already produced continue to show a waveform after the setting is turned off.

Show deactivated agents for last N days in Agent Filter — how long deactivated agents continue to appear in the agent filter after deactivation. Defaults to 60 days.

Use default SMTP server — when enabled, Ender Turing's built-in SMTP server is used and the custom-server fields are hidden. Turn this off to point the platform at your own SMTP server.

Host — the SMTP host name.

Port — the SMTP port (numeric).

Use TLS — enable to require TLS when connecting to the SMTP server. When Use TLS is on, Login must also be filled in: outgoing email is treated as not configured when TLS is enabled without an SMTP username, even if the form saves successfully.

Login — the SMTP username. Required when Use TLS is enabled.

Password — the SMTP password. When a password is already saved, this field shows a masked placeholder rather than the stored value, so the eye icon cannot reveal the real secret. The eye icon only toggles whether the text you have typed into the field is shown as plain text, which is useful when entering a new password. To rotate the password, type the new value over the masked placeholder before saving.

From Email — the address used in the From header of outgoing messages. Required when a custom server is configured and validated as an email address.

Changes you save in the General Settings card are saved separately from changes in the SMTP Settings card. Each card has its own Save changes button.

Most settings take effect immediately for new activity and are not applied retroactively — for example, turning Enable waveform generation on does not produce waveforms for audio that was already processed, and changing Convert spoken numbers in transcripts does not rewrite existing transcripts. Audio retention is the exception: see the next note.

Lowering the audio retention is destructive for already-stored audio. The scheduled cleanup reads the current Max days to store recognized audio files value each time it runs and deletes every non-playlist audio file older than that value. If you reduce the retention from 365 days to 30 days, audio recorded more than 30 days ago can be removed on the next normal cleanup. Conversation transcripts, review scores, feedback, and playlist audio are kept independently and are not affected. Confirm with stakeholders before lowering this value, especially in cloud (multi-tenant) deployments.

The audio retention field accepts whole days only. Fractional values such as 1.9 are rejected; values such as 365 and 365.0 are accepted.

The audio retention warning above 365 days only appears in cloud (multi-tenant) deployments. It is informational — it does not block saving.

In cloud (multi-tenant) deployments, fields that are read-only for your role still show the current value so you can confirm how the workspace is configured before asking for a change.