Smart features FAQs

Do we need to enable Smart everywhere at once?
No. Smart can be enabled at company level and then rolled out mailbox by mailbox, so teams can start with a small pilot and expand gradually.
What should we check before a pilot?
Confirm that Smart is enabled for the company, the pilot mailboxes are opted in where required, and body-ingestion participation is active for workflows that depend on Smart text outputs.
What is a sensible first pilot group?
Choose mailboxes with clear customer-facing traffic and managers who can review logs regularly. Avoid starting with unusually noisy or highly sensitive mailboxes unless that is the specific use case being validated.
Can different mailboxes have different Smart experiences?
Yes. Smart experiences can depend on company-level access, mailbox-level participation, body-ingestion participation, and which product surface is being used.
Will Google mailboxes need to be re-authenticated for Smart?
Existing Google mailboxes may need to be re-authenticated before email bodies can be ingested for Smart features, because body ingestion uses the Gmail read-only scope.
What is the difference between pausing Smart and deleting Smart data?
Pausing Smart is reversible and keeps existing Smart data until deletion is requested. Deleting all Smart data is permanent and removes stored Smart-derived insights and encrypted email body content.

Can Smart classifications work while message text is hidden?
Yes. Privacy mode can hide body text, Smart summaries, and evidence snippets while keeping Smart classifications visible where available.
Does privacy mode disable Smart?
No. Privacy mode changes what Smart text outputs are shown in the UI. It does not, by itself, disable Smart feature controls.
Does timetoreply redact PII before Smart enrichment?
No. We tested automated redaction, but it reduced result quality. Instead, exposure is reduced through opt-in body ingestion, mailbox-level participation, privacy mode, Amazon Bedrock inference, and encryption of stored email body content.
Is customer data used to train AI models?
No. Customer data is not used to train AI models.
Why do some conversations have more Smart fields than others?
Signal depth varies by message direction, available data, page context, and whether Smart enrichment has run for that conversation or message.

Do Smart filters replace existing message filters?
No. Smart category exclusions and Smart filters work alongside existing domain, email-address, and advanced message-filter rules.
What is the difference between Smart category exclusions and advanced Smart filters?
Category exclusions remove low-value traffic where global exclusion filters are applied. Advanced Smart filters help users narrow logs or reports by Smart dimensions such as category, needs response, sentiment, urgency, and intents.
Who can see Filtered Out logs?
Filtered Out mode is intended for company admins and shows excluded items with reason tags so filtering behavior can be reviewed and tuned.
Should managers use Filtered Out mode daily?
Usually no. It’s intended as an admin review and tuning tool, not a day-to-day manager queue.

When should we add Smart criteria to alerts or priority workflows?
Add them after logs show enough Smart signal quality for the intended use case. A short pilot helps managers understand what will be surfaced and why.
What happens when an alert email has incomplete Smart data?
When Smart is enabled but fields are unavailable, alert emails show a fallback Smart message rather than pretending values exist.
When do Smart columns appear in exports?
Smart columns are included when Smart access is enabled and body-ingestion opt-in is active. Exports continue to work without Smart columns when those conditions are not met.
Is the Manager Priority List available to everyone?
Not yet. It is in controlled rollout and access is limited while final validation completes.

Can users correct Smart values?
Yes, where the logs UI supports correction and the account has the required Smart/body-ingestion access. Corrections help create a human-in-the-loop review path.
Should one correction change every similar message automatically?
No. Treat corrections as review actions for the specific Smart value shown in logs unless the product explicitly says otherwise.

What security certifications does timetoreply maintain?
timetoreply maintains SOC 2 Type II and ISO 27001 certifications and undergoes independent security audits annually.
Who should customers contact for security or permission questions?
Contact your account manager or support@timetoreply.com for support, security questions, permission review, audit requests, or privacy-mode questions.