Importing existing records

Agents can import historical Tickets and inventory Assets from CSV. Employee accounts come from real Microsoft 365 sync or invitations; the production importer does not create Microsoft employees. Knowledge articles use their own Import articles workflow.

Prepare the file

Export a CSV with a header row and at least one data row. UTF-8 and Windows-1252 are accepted, with comma, semicolon or tab separators. The server accepts at most 50 MB per file: split larger exports into separate files. There is no fixed row-count limit within that size limit.

Review dates, identifiers and people before starting. Times lacking an explicit timezone are interpreted in your browser's displayed timezone and stored in UTC. Keep a source copy outside Empeiro for reconciliation.

Ticket fields available for mapping are User's Description of the Issue, IT Definition of the Issue, Resolution Details, System, Priority, Status, Assigned to, Date reported, Date completed, Issue logged by, Department and Title. This imports ticket records; it is not a general import of historical chat conversations or attachments.

Asset fields are Description, Asset tag, Site, Category, Assigned site, Assigned room, Assigned to, Status, Brand, Model, Cost, Purchase date, Serial number, Checked out on, Decommissioned date and Notes.

Choose and map a file

1. Open Import and select Tickets or Assets. Assets also offers Import a CSV.
2. Drop the file into the upload area or select Choose a CSV file.
3. Review the filename, row count, encoding, separator and warnings.
4. Under Columns, compare Your column and its sample Values in it with Goes into. Empeiro guesses from headers; select the correct destination or Don't import for each column. If multiple source columns map to the same field, only the first is used and a warning appears.
5. Review What will be written. The first twenty rows are displayed; counts cover the whole file. Distinguish Will be imported, Imported, flagged, Duplicate, skipped and Can't be imported. A flag may mean a value was left empty, an original value was retained in notes, a person was unmatched, or a new choice will be added. Read each reason.
6. To start over before importing, select Choose another file, or press Escape outside a dropdown.

The preview writes no ticket or asset. Uploaded content is kept in server memory for two hours and is lost on restart. If it expires, choose the file again. Nothing from the import file is sent to an AI model by the parsing process.

Understand duplicates and people

For tickets, a duplicate has the same Issue logged by, Date reported and first 80 description characters as an existing ticket or an earlier row. For assets, a matching Asset tag or Serial number in the inventory or earlier file rows causes a duplicate skip. A changed identifier can cause a record to appear new; inspect the preview rather than assuming repeated imports always skip everything.

People are matched against the organization's real directory. Resolve missing or ambiguous identities through the directory or account workflow, then preview again. Add people from this file to the simulated directory is a local-development control, not a production onboarding method. Do not use it for a firm's tenant.

The preview lists New values that will be added to your lists. Review these for spelling variants before confirming, because importing can extend classifications and inventory choices.

Optional assistant assignment

For tickets, Assign open tickets to the assistants is off by default. Leave it off for a historical import. Imported tickets are not automatically answered unless you enable this option.

When enabled, eligible open tickets with requesters in the directory are dispatched to Front line assistants, preferring their System specialties and otherwise in turn. Conversations open one a second and may hand off to IT Support. Completed tickets and unknown requesters are not assigned this way. This is real assistant work and can use the configured provider.

Import and reconcile

1. After reviewing mapping and counts, select Import [number] tickets or Import [number] assets. The server recalculates the plan against current records before writing.
2. Read progress. Larger imports show row counts and percentage; leaving the page does not stop the running import. Only one import for the organization may run at once.
3. Read the completion message, including imported, skipped, flagged and failed rows. A failure does not imply earlier writes were rolled back.
4. Select Download the report of flagged and skipped rows (CSV) when available. Expand The [number] rows in the report for row-level reasons.
5. Select Open the queue or Open the assets and compare representative records with the source, including people, dates, status and values. Ticket history keeps original dates for reports and records the import filename; assets receive a source-file-and-row history entry.
6. Select Import another file for the next file. Earlier tickets imports or Earlier assets imports lists saved run counts and outcomes.

Download detailed reports promptly: progress detail and report files are held until the server restarts. Earlier run summaries remain, but a summary is not the complete original export or row report.

Re-read previously imported tickets

Under Earlier tickets imports, select Re-read imported tickets beside a completed file. This reapplies current parsers and people matching to imported ticket values already held in the help desk; it does not reload the original CSV from your computer. Only fields whose parsed result differs are changed. The result lists ticket links and individual changes.

Use it to correct earlier exported-cell parsing, empty priorities or people who can now be matched. Review changed tickets afterward. It is not an undo, backup restore, or a way to re-import attachments. The local-development Add people from this file switch is unrelated to real employee provisioning.

Failure recovery