By collapsing code, logs, team knowledge, and past conversations into a single Cursor session, we've removed the context-gathering bottleneck for most of our support work.

Cursor

Cursor's technical support team shares how they use Cursor itself to handle over 75% of support interactions, achieving a 5–10x throughput increase. Their workflow centers on multi-root workspaces for cross-repo code investigation, MCP server integrations (Datadog, Slack, Notion, Linear) to pull customer data and logs directly into sessions, and parallel subagents (LogInvestigator, KnownIssueMiner, TicketWriter, CustomerReplyDrafter) that automate common investigation steps. The result is an AI-native support process that eliminates tool-switching and enables a small team to scale with a rapidly growing user base.

How technical support at Cursor uses Cursor · Cursor