We have shipped a big update that helps teams run AI Agents with more visibility and more control. The highlight is the new Analytics dashboard, designed to make it easier to understand what is happening across conversations, AI Agents, and Knowledge Bases.
Below is a recap of what is new and improved.
New Analytics Dashboard
The new Analytics area is now available in the dashboard. It is designed to help you track performance, spot issues earlier, and understand what is working across your workspace.
It includes four main tabs:
- Overview
A high level view of what is happening in your workspace, with the most important trends and operational signals. - Conversations
A deeper look at conversation activity and outcomes, useful for monitoring workload, assignment behavior, and what is happening in the inbox. - AI Agent
Agent level visibility to help you understand how AI Agents behave in production, how they perform, and where they may need tuning. - Knowledge
Knowledge Base performance and coverage, so you can quickly see what gets answered, what gets missed, and where to improve content.
If you want the full walkthrough, you can read it here: https://guide.tiledesk.com/manage-your-workspace/tiledesk-analytics
Knowledge Base improvements
We shipped several improvements to help you manage Knowledge Bases faster and monitor them more clearly:
- Knowledge Base KPI cards are now more interactive, with clickable functionality
- Added bar charts for answered and unanswered questions, plus line charts for answer rate
- Updated the format of Knowledge Base statistic counts
- Fixed an issue where Knowledge Base statistic links to the new Analytics did not work
- Default automatic content extraction for web content is now set to true
- Several Knowledge Base web content dialogs were refined to improve layout and navigation, including clearer scraping rule visibility and a smoother sitemap import experience
AI settings and model updates
- Improved AI settings dialog tooltips
- Added OpenAI GPT-5.5 model support
- Added HyDE and Use cache options in AI Settings for Ask Knowledge Base
- Fixed an issue where streaming preview was not re-enabled after disabling Use cache
- Fixed model display issues in Knowledge Base preview for Ollama and vLLM
- Fixed the default maximum token value for a template that was incorrectly too low




