The Dream of Autonomous Driving Needs Open Roads

Linear unveiled its Self-Driving SaaS vision, in which software operates autonomously on the user's behalf. This goes beyond simple assistive features, referring to an autonomous workflow system that applies Continuous AI to tasks such as extracting support tickets, analyzing issues, assigning teams, implementing code, and updating documentation.

Levels of Autonomy and Their Limits

Linear defined autonomy across three levels: Low (suggestion), Moderate (editable automation), and Full (fully autonomous). However, the key gap pointed out is that Linear cannot build every workflow that every team needs. Because each company has its own unique support systems, deployment procedures, and document repositories, simply waiting on the vendor's roadmap is inefficient.

Developer-Led Workflow Building

The competitive advantage lies in the workflows that Linear does not build. Therefore, developers should be able to define their own triggers (e.g., Sentry errors, Slack messages), compose chains, connect tools (such as MCP servers), and measure outcomes based on business metrics. As observed by the Continue team, team-specific workflows connected to external tools like PostHog, Sentry, and GitHub are being implemented without waiting on Linear's roadmap.

Ultimately, self-driving SaaS isn't something you buy—it's something developers build themselves, tailored to how they actually work.