Book a demo

By continuing, you're agreeing to SparrowDeskPrivacy PolicyandTerms of Service

Sign up modal image

Sign up

Get started with your 14-day free trial.

By continuing, you're agreeing to SparrowDeskPrivacy PolicyandTerms of Service

Sign up modal image

Changelog

Stay updated with the latest features, improvements, and fixes in SparrowDesk.

February 28, 2026

Download Reports as PDF

Download Reports as PDF

You can now download reports as PDF files, making it easier to save, share, and review report data outside SparrowDesk.


The exported file captures your current view, including applied filters.


Use this to share weekly performance updates with leadership, download SLA reports for audits and reviews, or save report snapshots for internal reporting.


Along with this ,


Reports for CSAT - You can now track customer satisfaction with a dedicated CSAT report. Use it to review feedback trends, measure satisfaction over time, and understand how your team is performing.

Optimized contact and ticket deletion to reduce delays and improve performance


Upgraded routing and asset loading for better application stability


Removed email notifications for self status changes

Fixed report filters and custom date ranges showing incorrect or missing dates


Fixed report charts showing inaccurate dates or incomplete data in some views


Fixed drill-down issues in reports when switching views or navigating across modules


Fixed filters like Team, Merge, and Requested By not working correctly in reports


Fixed CSAT not triggering when live chats were closed using macros


Fixed image previews not appearing for some customer attachments


Fixed an issue where previously uploaded content could be attached to a new conversation

February 20, 2026

AI Classification in workflows

ai-classification-in-workflows

Bring AI-powered decisioning into workflow automation.

AI Classification introduces an AI-native workflow action that replaces complex keyword-based if/else logic with structured, deterministic classification. Admins can define a short prompt and a fixed set of buckets, allowing AI to always select exactly one predefined outcome and drive workflow branching with semantic context.

This action can be triggered on new conversations or incoming messages, reads the latest inbound message, and routes workflows based on intent, sentiment, language, or other defined criteria. Classification is constrained to predefined buckets only, ensuring workflows remain predictable and never fail due to AI output.

Available on the Professional plan in v1.

Why this matters

  • More precise automation – Route conversations based on meaning, not just keywords.
  • Predictable outcomes – AI always resolves into predefined buckets for reliable workflow branching.
  • Reduced workflow complexity – Replace layered if/else rules with a single AI-powered decision step.
  • Scalable support operations – Handle nuanced routing, SLAs, and assignments with semantic understanding.

Watchers for tickets

Keep the right people in the loop on every conversation.

Agents and managers can now watch tickets to stay informed on important updates. Any agent with access to a ticket can follow it, while managers and admins can add or remove watchers manually or through workflows to ensure the right stakeholders are included.

Watchers receive email notifications when key activity happens — including customer replies, agent responses, internal notes, and status changes — and can view all followed tickets in a dedicated Watching view within the inbox. Notifications are automatically suppressed for actions triggered by the watcher to avoid noise.

Admins can also add watchers automatically using workflow actions, making it easier to loop in creators, owners, or specific team members when a ticket is created.

Why this matters

  • Better visibility – Keep stakeholders informed without manual updates.
  • Faster collaboration – Ensure the right teams stay aligned on critical tickets.
  • Less noise – Receive updates only on conversations that matter.
  • Scalable workflows – Automatically add relevant watchers as tickets are created or updated.

AI support for image attachments in conversations

AI can now process and analyze image attachments shared in email and chat. By understanding visual context alongside messages, the system can generate more accurate responses and improve overall conversation handling.

  • Resolved a crash on the Roles page when accessing roles configured with ticket and form permissions.
  • Assignees now receive notifications when internal notes are added to their tickets.
  • SLA reminders are now correctly delivered to the requestor’s inbox.
  • Portal notification customization has been removed from the Notifications section, with updated permission labels and naming.
  • Minor UI improvements on the Email configuration page.
  • Articles created by deleted members now open without errors.
  • Zoona replies no longer appear as edited in AI Agent Improvements.
  • Eliminated unintended horizontal scrolling across affected pages.
  • Corrected preview naming inconsistency where Luna appeared instead of Zoona.

February 14, 2026

Publish Help Center articles as Unlisted

Publish-as-Unlisted

Help Center articles can now be published as Unlisted with a shareable preview links.Unlisted articles remain hidden from collections and public search, but anyone with the link can access and review them. This enables teams to collaborate on drafts, gather approvals, and share sensitive content without exposing it before publication.


  • While writing Help Articles, authors can now link directly to specific sections within an article using anchor links and quickly connect to other Help Center articles while writing.
  • Fixed an issue where newly added conversation dropdown options appeared duplicated after saving.
  • Restored password login visibility for account owners and admins when SSO is enforced on domain-specific login links.
  • Resolved inconsistencies affecting login functionality when SSO is enabled for an account.

Added a public API to retrieve all tags in an account, enabling programmatic access to tag data for integrations, reporting, and automation.

October 13, 2025

Create workflows that trigger when multiple fields are updated in a conversation

multi-field-update-trigger

With this update, SparrowDesk now allows workflow automations to trigger when multiple conversation fields are updated, using the Conversation updated trigger.

Admins and agents can specify multiple fields as trigger conditions and define whether the automation should run when all conditions are met (AND) or when any condition is met (OR). This gives teams greater control over when workflows fire and helps avoid unnecessary or premature automation.

Why this matters

  • More precise automation – Trigger workflows only when the right combination of fields changes.
  • Reduced noise – Prevent automations from running on partial or irrelevant updates.
  • Greater flexibility – Support complex use cases without creating multiple workflows.
  • Better alignment with real processes – Automations can now mirror how conversations actually evolve.

This update makes workflow automation more powerful and predictable, enabling teams to build smarter, context-aware processes at scale.

  • Optimized AI-based chat subject generation to produce clearer subjects in a single message.
  • Fixed an issue that prevented email content from rendering correctly and caused screen overflow.
  • Resolved macro assignment failures that could disrupt ticket workflows.
  • Fixed a macro-related issue that caused activity logs to crash in certain scenarios.
  • Resolved an issue that prevented workflow triggers from being configured for unassigned conversations.

Restricted AI Agent configuration APIs to admin-level users to improve access control and configuration safety.

October 10, 2025

Use AI to generate the subject line for conversations

AI-based-chat-transcript

Support teams often create tickets on behalf of customers during conversations. With this update, SparrowDesk now automatically generates a relevant ticket subject using AI when an agent records an issue described by the customer.

The AI analyzes the conversation context at the time of ticket creation and generates a concise, meaningful subject that reflects the customer’s reported issue. This removes the need for agents to manually summarize or name tickets, while keeping records clear and consistent.

Why this matters

  • Clear context at creation – Tickets are accurately titled the moment they’re logged.
  • Reduced manual effort – Agents don’t have to pause and write subjects themselves.
  • Better ticket organization – Consistent, descriptive subjects improve scanning and filtering.
  • Smoother handoffs – Automatically generated subjects preserve context during reassignment or escalation.
  • Improved reporting – Well-structured subjects make issue trends easier to analyze.

This helps teams capture customer-reported issues more efficiently, keeping the ticket inbox organized and scalable as volumes grow.

  • Added support for email-based filtering of contacts and requestors.
  • Fixed an issue where chat conversation subjects did not display the requester’s full name.
  • Resolved an issue preventing source links from opening in AI-suggested replies during live chats.
  • Fixed a permission issue where agents could access tickets assigned to other users.
  • Resolved a visibility issue where team-assigned tickets were visible to non-team members.
  • Fixed workflow time input to prevent decimal values and ensure valid configurations.