How Asana automated
70% of its internal support

At Asana, the IT team received more than 400 internal tickets per week: password resets, software access requests, questions about internal tools. These mostly repetitive requests tied up 3 full-time engineers and slowed down the resolution of critical incidents.

Leadership was looking for a solution able to handle these level-1 tickets autonomously, without custom development or an overhaul of the existing helpdesk system.

In 18 days, the Asana team deployed an EzyAgents AI agent connected to their Zendesk instance. The agent was trained on the internal knowledge base (Confluence, Notion), IT procedures and 6 months of resolved ticket history.

• Automatic analysis of every new ticket, classified by priority and type
• Autonomous replies to level-1 requests (resets, access, FAQ)
• Escalation to a human engineer with full context for complex tickets
• Mandatory HITL approval before any action on production systems
• Automatic weekly report on volumes, types and resolution times

From the very first week in production, the agent handled 52% of incoming tickets without human intervention. By the end of the first month, that rate reached 70%. IT engineers cut the time spent on level-1 support by 85%, letting them focus on higher-value infrastructure projects.

Cost per ticket dropped from €18 to €7, a 3× ROI over 5 months including the EzyAgents subscription. The internal satisfaction score (measured quarterly) rose by 45 points.