📌 Note: This article accompanies the Element451 + Slate resource set. If you have not set up the integration yet, start with Using Element451 Bolt with Technolutions Slate (Overview).

No. This integration is a scheduled file sync, not a live API connection. Bolt agents personalize from the data most recently synced into Element451, which is as fresh as your import schedule (daily for most schools). If a record changed in Slate this afternoon, the agent will know tomorrow. The scheduled import/export approach is the supported integration method for Slate, and the lever for freshness is sync frequency. If you need fresher data, schedule more frequent syncs; your Element451 team can help you tune the cadence.

Yes. Conversation summaries and full conversation transcripts can both sync to Slate as Interactions. Add the Full Conversation Transcript field to your conversations export and map it in your Source Format, as shown in the outbound walkthrough. A transcript is created when a conversation closes, so a conversation that is still open exports without one. Keep the topic or summary in Interaction Subject so each conversation stays legible on the Slate timeline.

Slate decides, per record, whether to create, update, or ignore based on its own matching rules, and only the fields your Source Format maps are written. Nothing else is touched. You stay in control on both sides: in Slate via Source Format mappings and the Update-Only setting, in Element451 via import mode (Update Existing and Import Only New People is the usual choice). We recommend dedicated Slate fields for Bolt engagement data so AI outputs never collide with fields owned by other processes.

Element451 exports typically include all contacts in the selected segment during each run, not just changes. Slate's matching rules sort out what is new versus an update, which is what lets updates to existing contacts flow automatically.

Only identified visitors become records. A Bolt agent captures identity through the Inquiry Flow skill (first name, last name, email, academic interest), so confirm it is enabled on each student-facing agent. Anonymous visitors who never share contact information do not create records and cannot be exported. See Bolt Agent Skills.

Yes, but treat column changes as a coordinated change on both sides. Adding a column to a Slate export means updating the Element451 import mapping, and vice versa. Removing or renaming a column without updating the other side is the most common way these integrations break, and it can break silently (see Troubleshooting below).

Yes. Build and validate the full round trip against your Slate test environment with a small batch of test records (2 to 3 is plenty for the first pass), then repoint the SFTP connections and Import Automations to production. This is how the reference implementation was built and it caught nearly every issue listed below before students were involved.

Check the file header row first. A missing or renamed header column can break the mapping without raising an obvious error, and the symptom is simply that nothing imports. Compare the file's header row against the Source Format mapping column by column. In one real case, a single missing header column silently stopped the entire conversation feed.

The summary is probably mapped only to Interaction Comments. Map the conversation topic (or the summary itself) into Interaction Subject. Comments-only Interactions may not appear in the interaction list view on the person record. See the outbound walkthrough for the full mapping table and display format options.

A transcript is created when a conversation closes, so a conversation that was still open when the export ran has no transcript yet. This shows up most in the export Preview tab for students with several open conversations. Check whether those conversations have been closed. Include Conversation GUID in the conversations export so each row shows which conversation it belongs to, and add Conversation Status if you want the file to show whether each conversation has closed. If a conversation is closed and its transcript is still empty, contact Element451 support with its Conversation GUID so the team can check it.

Confirm the leads file includes birth date. Slate's dedupe uses first + last + email + DOB, and without DOB it has less to work with. Also confirm the Element ID field is mapped, populated, and used for matching, ideally flagged as Unique for Merging (see the identifier article).

Confirm the inbound file carries a stable unique identifier (email, Slate ID, or an external ID) and that your import's Matching settings use it. Watch for populations where the same person appears under different email addresses.

Almost always import order. The lead import must process before the conversation import so person records exist (with Element ID populated) before Interactions try to attach. Configure the order in Slate Import Automation, stagger the Element451 export schedules so the leads file lands first, and verify processing order in Slate's Job Activity Monitor.

If two Slate records merge and the surviving record's identifiers change, incoming Element451 data can lose its match. Flagging the Element ID field as Unique for Merging lets Slate keep matching on it even after merges. See Technolutions: Unique for Merging.

Check credential expiration first, on whichever side owns the failing connection. For the inbound feed, that is the Element451 SFTP credentials stored in Slate's scheduled export connection. For the outbound feed, that is the Slate SFTP credentials stored in the Element451 export destination. Update the stored credentials after any rotation. Also confirm nothing changed in IP allowlisting. Slate's outbound IPs are published in Technolutions' Outbound Networks.

Schedule imports 15 to 30 minutes after the export is due to land, and toggle on When no matching files are available, execute the next scheduled task run so a single miss does not halt the schedule. See Scheduling Import + Export.

Open Run History and read the skip reasons. The common causes are matching conflicts (two fields matching two different profiles, skipped by design) and rows failing a validation. See Creating Imports for matching behavior.

Still stuck? Contact Element451 support with the file name, the run timestamp from Run History, and the Job Activity Monitor status. Those three things shortcut most investigations.