What Is Control Desk and How To Use It
Once your AI agent is live, the Control Desk becomes your mission control. It’s where you monitor performance, fine-tune behavior, review conversations, and manage all key settings, without needing to reconfigure from scratch. Here’s how to make the most of it.
Overview: What the Control Desk Offers
The Control Desk is the central dashboard of each AI agent project. It provides a real-time snapshot of:
- Current integrations with helpdesks.
- Connected knowledge bases.
- A playground to test your AI agent’s responses.
- Easy access to analytics, configuration, and version control.
This centralized view is designed to make support management more agile and transparent, whether you’re running one project or many.
Managing Helpdesk Integrations
In the Integration Status section, you can:
- See which helpdesk (e.g. Zendesk, Intercom, Freshdesk) is currently connected.
- Modify your integration settings.
- Remove an integration and switch to another, without rebuilding the AI agent.
Note: This flexibility allows you to migrate platforms while preserving all your AI setup and training.
Connecting and Organizing Knowledge Bases
Directly beneath the integration section is Data Sources. From here, you can:
- Connect or disconnect specific knowledge bases with one click.
- Assign multiple knowledge bases to the same AI agent (or just one, if preferred).
- Maintain clean organization and modularity across projects.
Every change takes immediate effect without requiring retraining or resetting the AI agent.
Testing with the Playground
On the right side of the Control Desk is the Playground, a real-time testing environment. Use it to:
- Simulate user questions and preview answers.
- See which data sources were used to generate each response.
- Expand the test window and refresh the chat anytime.
It’s perfect for QA, training review, and content verification before going live.
Project Settings & Rate Limits
Click the Edit button (top right) to:
Rename your project
Set rate limits per project:
- Max number of responses per session.
- Max number of conversations per day.
This helps control cost and performance while scaling safely.
AI Analytics Dashboard
Under the Analytics tab, you’ll find detailed performance metrics, including:
- Resolution rate: % of conversations fully handled by the AI.
- Total resolved, escalated, and in-progress sessions.
- First response time and average resolution time.
- Customizable time ranges and trend views.
Use these metrics to identify success gaps, enhance training, and track ROI.
Reviewing Conversations
The Conversations tab is your AI audit trail. For every session handled:
View the full transcript
See which data source powered each answer
Filter by:
- Status (resolved, escalated, in progress).
- Timeframe (last 7 days, custom range, etc.).
Use case tip: Filter for escalated conversations → identify gaps → update the knowledge base with FAQs or new documents.
AI Configuration & Versioning
Under the Configuration tab, manage:
- AI persona and tone of voice.
- Response style settings (length, emoji use, etc.).
- Language and keyword restrictions.
Edits generate a new version, which doesn’t go live until you manually deploy it, giving you time to test safely. You can view and switch between multiple versions, making it easy to roll back to a stable release if needed.
Why It Matters
The Control Desk is more than just a settings page, it’s your command center for support automation. With CoSupport AI, you’re not just launching a chatbot. You’re deploying a customizable, measurable, and enterprise-ready agent that evolves with your business.
