Skip to main content

Full Conversation Transcript Exports | October 2026

Full Conversation Transcript is now a separate field in conversation exports and the export API, so Slate and other systems of record can receive the complete exchange alongside or instead of AI Summary

K
Written by Kelly Milnes

Released: October 2026

Overview Header

Until now, conversation exports carried a conversation's metadata and its AI Summary. Teams working in Slate or another system of record could see that a conversation happened and what it was about, but not what was actually said.

Full Conversation Transcript closes that gap. It is a new field that carries the complete, readable exchange between the student and your Bolt Agent or staff, available in scheduled and manual conversation exports and through the export API. AI Summary stays exactly where it is, and you choose whether to send the summary, the transcript, or both.

Details Header

One transcript, built once and reused everywhere

When a conversation closes, Element451 now builds a single plain-text transcript and stores it on the conversation record. Every export and API request reads that same stored value, so Slate, an SIS, another CRM, or a custom integration all receive identical content for the same conversation.

Each transcript includes

  • Every readable message, in chronological order

  • A speaker label on each entry, so it is clear who said what

  • A timestamp on each entry when one is available

  • Web chat, email, SMS, WhatsApp, and phone or voice content, including call transcriptions when a recording was transcribed

  • Attachment names and links, with HTML converted to clean readable text

Internal notes, prompts, hidden system and tool content, and moderation metadata are never included. The transcript only contains what the student could see or hear.

A separate field in your conversation exports

Full Conversation Transcript appears as its own field in Import + Export, grouped with the other conversation fields and separate from AI Summary. Add it to any scheduled or manual export the same way you add any other conversation field.

To find it, go to Data & Automations, open Import + Export, open the export task or mapping template you want to change, choose Mapping, then Add a Mapping and search for Full Conversation Transcript.

Existing exports and mapping templates are unchanged until you add the new field. When you add it, we recommend adding Conversation GUID too. The GUID identifies each conversation, so every row shows which conversation it belongs to even when its transcript is empty, and downstream systems can match each transcript to the right record. Adding Conversation Status also shows whether each conversation has been closed.

We recommend adding the following in your mapping:

  1. Full Conversation Transcript — Required to export the full conversation

    Sends the complete readable exchange, including available speaker names, timestamps, and message text.
    ​

    If omitted: The export will not include the full transcript. Other selected fields, including AI Summary, will still export.
    ​

  2. Conversation AI Summary — Optional

    Gives staff a quick overview without needing to read the entire transcript.
    ​

    If omitted: There will be no summary column. The full transcript is unaffected.
    ​

  3. Conversation GUID — Recommended

    Uniquely identifies each conversation, making it easier to trace an exported row back to Element451, troubleshoot issues, and reconcile repeated exports.
    ​

    If omitted: It becomes harder to identify which conversation a row represents or recognize the same conversation across multiple exports. The GUID identifies the conversation, not the student.
    ​

  4. Conversation External User Id — Recommended when used as your person-matching key

    Identifies the student or contact associated with the conversation in Element451. Your receiving system must recognize this ID, or you must export another agreed person-matching identifier. This is not automatically a Slate ID or SIS ID.
    ​

    If omitted: The export still works, but without another suitable person identifier, your receiving system cannot reliably attach the conversation to the correct student. For Slate, our documented setup uses the contact’s Element ID, matched to the Element ID field stored in Slate.
    ​

  5. Conversation Status — Recommended for validation and troubleshooting

    Shows whether a conversation is open or closed, helping explain why a transcript may be empty.
    ​

    If omitted: You will not have a separate status column to check. Omitting this field does not change which conversations are exported or whether their transcripts exist.
    ​

  6. Conversation Delivery Method — Recommended for channel context

    Shows whether the conversation happened through email, SMS, phone, WhatsApp, or chat. This is useful for reporting and mapping conversations to interaction types in your receiving system.
    ​

    If omitted: The export will not include a separate channel indicator. If your receiving system requires one, it will need another source or a configured default.
    ​

  7. Conversation Created At — Recommended

    Provides the conversation’s creation timestamp for chronological reporting and activity-date mapping.
    ​

    If omitted: The export will not include a separate conversation creation date. If your receiving system requires an activity date, it will need another date source or a configured default.

Available through the export API

Integrations that already pull AI Summary through the export API can now pull the full transcript in the same request. Add the conversations-full-transcript mapping (Full Conversation Transcript in the export mapping picker) as a column in your POST /v2/users/export template, and unwind on conversations-root so each row is one conversation. It returns the complete plain-text transcript, the same text a file export contains. Credentials, limits and parameters are the same as for any other export.

We recommend adding conversations-guid and conversations-status too, so each row shows which conversation it is and whether it has closed. Here is an example request body with all three. It returns one row per conversation for the students in a segment.

{
"item": {
"options": {
"column_key": "slug",
"unwind": { "root": "conversations-root" }
},
"template": {
"columns": [
{
"field": "Element ID",
"mode": "slug",
"slug": "user-elementid"
},
{
"field": "Conversation GUID",
"mode": "slug",
"slug": "conversations-guid"
},
{
"field": "Conversation Status",
"mode": "slug",
"slug": "conversations-status"
},
{
"field": "Full Conversation Transcript",
"mode": "slug",
"slug": "conversations-full-transcript"
}
]
},
"segment": "YOUR_SEGMENT_GUID"
}
}

Replace YOUR_SEGMENT_GUID with the guid from your segment's URL, or send a users list instead. The headers and a sample response are in Example 5 of Exporting User Data. The export endpoint is available in V2 only for now; full endpoint details are in the V2 API documentation.

Transcripts are opt-in. Summary-only integrations keep working exactly as before and never receive a large transcript payload they did not ask for. Identifiers, response structure, and pagination are unchanged.

This is a good fit for teams that run their own nightly analytics or reporting on verbatim conversation text.

Voice Conversations

Completed phone calls are now closed automatically once the call ends, which means voice conversations receive a transcript the same way chat, SMS, and email conversations do.

Benefit Header

Why teams asked for this

This release comes directly from institutions that told us the AI Summary alone was not enough. The reasons varied, but the request was the same.

  • Advisors working in Slate wanted the full exchange on the student record so they could continue the conversation without switching systems

  • Compliance and student conduct teams need the complete conversation, not a summary, to meet legal and documentation requirements

  • Analytics teams want verbatim call and chat text for their own nightly reporting and quality review

  • Communications teams want a complete SMS history they can export for a given date range

Good to know

  • A transcript is created when a conversation closes, so open conversations do not have one yet. If Full Conversation Transcript is empty in the export Preview tab or in an exported file, check whether that conversation has been closed. This is most noticeable for students with several open conversations, and including Conversation GUID in your mapping shows which conversation each row belongs to

  • If a conversation is closed and its transcript is still empty, contact Element451 support with its Conversation GUID so the team can check it

  • If a conversation is reopened, exports keep showing its earlier transcript until it closes again. Once it closes, the transcript is rebuilt to include the new messages

  • Very long conversations are subject to a size limit. A transcript that reaches the limit ends with a note saying it was cut short, so partial text is never presented as complete

  • This release does not add a new transcript viewer inside Element. Staff continue to read conversations in the Conversations view as they do today

  • Slate schools can map the transcript to Slate Interactions by following Element451 to Slate: Syncing Bolt Outputs Back to Your CRM

Support Docs Header

Use these guides to add the transcript to your exports and send it to Slate

Did this answer your question?