Skip to main content
LLMs.md

Approval & Customer Sending

Submitted form reports can be sent to the customer directly — or only after an explicit approval. This page covers both steps: the approval loop and the send dialog with AI draft.

Approval before customer sending​

The toggle on the template​

Whether a form goes through the approval loop is decided by the template: in the form editor, the template properties include the toggle "Approval before customer sending". When it is on, submitted responses must be approved or ignored by an admin before they go to the customer. When it is off (default), everything behaves as before — sending is available immediately.

Queue in the reports list​

Submitted responses of an approval-required template carry the badge "Approval pending" in the reports list. In the header filter row, the Approval filter narrows the list down to exactly these open decisions — so you can work through the queue deliberately.

Approve or Ignore​

In the report detail, a waiting response offers two buttons:

  • Approve — the report is cleared for customer sending; the detail view then shows "Approved on" with the date.
  • Ignore — the report does not need to be sent to the customer (e.g. an internal test run or a duplicate submission). It remains fully stored and visible, but counts as decided.

The decision always applies to the response as a whole, never to individual fields. Both actions also work on the go: in the LiteLog app, the report detail shows the same Approve and Ignore buttons.

Automation​

Under Automation, the triggers "Report approved" and "Report ignored" are available. They let you, for example, automatically start the customer sending or notify the site manager as soon as a decision has been made. As always: without a rule that has been created and switched on, nothing happens on its own.

Customer sending with AI draft​

After approval — or directly, for templates without an approval requirement — you open the sending flow from the report detail. If the AI function "Translate forms & reports" is active, the send dialog opens with a ready-made AI draft:

  • Subject and cover letter are pre-written — in the customer's language, not yours.
  • Both fields are freely editable; your manual input is kept even if the draft is reloaded.
  • The option "Send polished version as PDF" attaches a refined PDF version: free-text answers are translated and linguistically polished. The PDF carries an AI notice footnote so the recipient can recognise the polished version as such — the original remains unchanged in the portal.

Without the AI function enabled, sending works perfectly well manually: you write subject and message yourself and send the report as usual. If the AI draft is ever unavailable, the dialog says so openly — "AI draft not available" — and lets you continue manually.

Sending from the LiteLog app​

The LiteLog app also offers the "Send to customer" button in the report detail. It opens the same sending flow with subject, message and the option "Send polished version" — so, for example, the site manager can send a handover protocol right on site.

Sent on​

After sending, the report detail — in the portal as well as in the app — shows the line "Sent on" with the date. That way you can see whether and when a report reached the customer without digging through a mailbox.

Good to know​

:::tip Note

  • The approval loop works without AI: the "Approval before customer sending" toggle and the Approve/Ignore buttons are part of the normal report flow. The AI only contributes the draft and the polished version.
  • Ignoring deletes nothing — the report remains in the list and in evaluations.
  • The recipient of the sending is derived from the customer of the property; the AI draft follows that customer's stored language. :::

:::info Data protection The AI function "Translate forms & reports" is optional and disabled by default. When active, processing runs via Azure OpenAI. The original of every response always remains stored unchanged, and AI-generated versions are labelled — in the portal via the "Translated / Original" switch, in the sent PDF via the AI notice footnote. :::