Import data & schedule syncs
Choose a source, confirm its destination, and monitor imports and scheduled updates.
Choose an import source
Open Data Import → Import New Data. The wizard follows Source → Destination → Sync → Review.
The wizard presents four standard source choices:
| Source | What to provide |
|---|---|
| My device | Drag files into the picker or browse your device. Stay within the 2 GB total size limit and the maximum file count shown for your account. |
| Microsoft SharePoint | Site URL and library or folder path, when this optional source is configured for your instance. |
| Google Cloud Storage (GCS) | A bucket, optional prefix, and the requested service-account JSON credentials. |
| Google Drive | Folder ID or URL, when this optional source is configured for your instance. |
SharePoint, Google Cloud Storage, and Google Drive are optional import sources configured when your Talvi instance was created. The available sources depend on your organisation’s setup. Custom import sources can also be added and appear here when configured. Their fields and connection requirements depend on the integration.
LMS and custom content integrations
As part of onboarding, Talvi works with your team to connect the LMS and other systems where your content lives. A custom LMS, proprietary content store, or internal API may require custom development work before it can appear as an import source. Share the source-system details with your Talvi team so access, implementation, and support can be agreed.
For each integration, agree the content and folders in scope, the authorised source account, the destination dataset, sync frequency, and how changes or removals are handled. Check which source permissions are carried into Talvi and define the agent audience accordingly; do not assume every upstream per-file permission is automatically inherited.
Support for a file format such as SCORM means Talvi can process supported packages. It does not by itself connect the LMS or establish enrolment, bookmarking, or completion writeback. Include any required learning-system workflow in the custom implementation scope.
See Talvi system overview and Security and Governance when planning the connection.
For accepted formats, see supported file types. A supported upload format does not necessarily have an inline preview.
File limits for device uploads
Each upload has a 2 GB total size limit and a maximum number of files that depends on your account. The upload area shows your applicable file-count limit. Both limits must be met, whether you select files together or add them through several browse or drag-and-drop actions.
Each selected source file counts as one file, including a SCORM ZIP or CPT archive. Once the maximum is reached, additional files are not added. Remove a selected file to make room, or wait until the current import finishes before starting another upload.
When another import or dataset creation is running
New upload controls are disabled while a current Data Import or Create Dataset operation is processing. Hover over a disabled upload control to see: “Please wait until the current Create Dataset or Data Import has finished.”
Wait for processing to finish, including any dataset preparation after the file transfer. Talvi checks the status automatically and re-enables uploads when processing ends. Follow progress in Active imports or Create Dataset → Dataset activity.
Confirm the destination
Select the dataset that should receive the files and add a short description. Check AI agents using this Dataset before continuing: new knowledge can affect more than one agent when they share that dataset.
If you need a new dataset, see Create Dataset. For an existing collection, you can inspect its current files in File Management.
Choose a sync schedule
Local uploads are one-off imports. Connected sources can offer recurring sync schedules, such as hourly, daily, weekly, or monthly. Choose No schedule for a one-off connected import. Use the options actually displayed for your source.
Review and start
Check the source, destination, description, and schedule on Review. Tick I’m ready to start the import, then select Begin data import. Large files can take time to upload and process; keep the upload page open until the transfer finishes.
Monitor and maintain imports
- Active imports shows running jobs and refreshes automatically. Open a job to inspect its details.
- Completed / Failed provides history, date filters, and a Refresh action. Inspect a failed job’s error before retrying.
- Scheduled data imports lists recurring connections. Use the available edit, pause/resume, run, or removal controls to maintain each schedule.
After a successful import, verify the files in File Management and try a representative question in Get Answers.
Follow the complete recipe: Update your AI agent’s data.
