The assurance system of action for auditors, consultants, and clients. Purpose-built for requesting and testing audit evidence against assertions and risks. Designed for the agentive AI era.
Contact Punchcard for current plans, volume pricing, and enterprise terms.
Trust profile
Punchcard introduced a structured triage interview process within CoAudit for reviewing automation results, allowing auditors to address findings through an interactive sequence. The update adds AI-written cell value previews before confirmation, a toggle for controlling AI citation behavior, and usability improvements including responsive Library loading and various bug fixes for editor stability and interface consistency.
Punchcard released several feature updates to its CoAudit product, including the ability to generate Excel workbooks with clickable citations and expanded delegation capabilities for complex tasks. Users can now upload folders and ZIP archives directly into the chat interface for processing. Additionally, the update adds support for filtering workpaper rows by review state, improves automation re-run logic, and upgrades the underlying AI model for enhanced accuracy.
Punchcard updated the CoAudit chat interface to provide consistent, readable activity logs for specialized helper agents. The update also resolves issues related to Excel exports containing cited workpaper cell references, ensuring support for large-scale workpaper data without re-reading source documents.
Punchcard now supports sort-aware workpaper exports, ensuring downloaded Excel files match the grid display order. The update also introduces clickable source chips within the cell detail panel to trace data origins and adds visual indicators for skill mentions in chat. Several bug fixes were implemented to improve the stability of bulk file uploads and engagement variable creation.
Punchcard has migrated its Deep Analysis functionality into the Skills library, allowing for direct import of existing questions. The update also addresses workpaper reliability issues, specifically improving state stability during task switching, ensuring that review status changes reflect across linked items, and implementing more robust error handling for failed saves.