Import People data from CSV and ask AI about the document on screen

Create employees, positions, and teams from a reviewed CSV instead of adding every record by hand. You can also open the AI Assistant from a process model or task instruction and ask about the document already on screen. Create employees, positions, and teams from CSV Select Bulk upload on the Employees, Positions, or Teams page to download a template or sample, upload up to 500 rows, check how your columns map to Klarify, and review the valid, existing, and invalid records before importing. Existing records are ignored rather than updated or duplicated, making corrected files safe to run again. The importer also detects common file problems before they reach review. It can recognize alternate separators and a title row above the headers, while files that are actually Excel, Numbers, OpenDocument, or PDF content receive a clear export-to-CSV message. Learn more: Import people from CSV. Import CSV files with Klarify AI The in-app AI Assistant now has a Download a CSV import template button beside the attachment button. Attach a completed employee, position, or team template and Klarify takes it directly to a preview and approval card without asking the language model to recreate the file. Counts distinguish records that were created, already existed, or failed. Connected AI clients can also import regions, locations, departments, assignments, and team memberships using human-readable names and emails instead of internal database IDs. Learn more: CSV bulk import. Ask the AI Assistant about the open document The AI Assistant is now available in the header of process models and task instructions. It receives the open document and its visible draft or published version as context, so you can ask “summarize this process,” “who owns each step?”, or “what does this task link to?” without finding an ID or re-entering the document name. Process answers can resolve performer candidates to the employees who currently fill those positions, identify vacancies, and show owner job titles. Suggested follow-up questions make it easier to explore the document, and each document keeps its conversation when you leave and return. Learn more: Use the AI Assistant on a document. Use a simpler, more dependable Klarify AI connection Klarify AI now presents 11 workflow-focused tools instead of dozens of individual record actions. This keeps the complete tool set within enterprise-client limits and gives Claude, Microsoft Copilot Studio, and other MCP clients clearer choices for searching documents, reading processes, finding people, and managing organization data. Existing integrations can continue calling older individual tools during the migration period. The same update improves the answers produced from Klarify data: Process summaries explain purpose, scope, boundaries, and participants instead of repeating every stage name. Questions about how work is done use approved, published content by default. An exact-name draft request still works, but the answer identifies it as an unpublished draft. Accepted invitations no longer appear as expired invitations. Organization authentication settings are offered only to roles that can actually read them. Documents directly inside a requested folder now include their name, type, and publication status. Learn more: Klarify AI tool reference. Move remaining legacy API integrations to scoped keys Organizations that still have a legacy read-only API key can continue viewing and copying it while they migrate. Klarify no longer creates or regenerates legacy keys, and organizations without one no longer see the legacy settings page. New integrations should use scoped, revocable API keys with bearer authentication. Learn more: Migrate a legacy API key.