Doctolib Transforms Development Practices with AI Agents

Starting January 2026, Doctolib has transitioned its development practices to be entirely agent-based, involving all 600 developers in this shift. Following initial experiments in 2025, the company adopted the ‘agentic-first’ approach, encouraging engineers to prioritize AI contributions in their workflows. This model supports programming strategies centered around specifications, thereby redefining the developer's role to that of an orchestrator ensuring quality and direction in the coding process. The integration of automation in managing documentation through tools like Open Spec emphasizes Doctolib's commitment to agile practices and adaptability in an evolving tech landscape.

The implications of this transformation are profound as Doctolib moves toward fully automated development pipelines. By employing a range of specialized agents for tasks such as ticket processing and production error management, the company is streamlining coding and review processes to minimize human intervention. This shift not only enhances the productivity of current development teams but also prepares the firm for scalability, with a greater emphasis on non-developer contributions towards development practices. As AI becomes a core component in its operations, Doctolib's strategy signifies a potential decrease in reliance on external AI tools, paving the way for greater national control over technological innovation.