Anthropic

Anthropic is an AI safety company and the developer of the Claude family of models, with an applied AI team that builds tooling and conducts research on agent workflows and long-horizon tasks.

Organizational culture

Anthropic has a distinctive onboarding practice described by one of its engineers: new hires are expected not to ask questions, instead figuring things out independently — mirroring the way an AI agent must operate autonomously. (Reflecting on a year of Claude Code, 20:04)

Research and product focus

Anthropic's applied AI team works on agent tooling, including proactive agent workflows built with Claude Code. (Build a proactive agent workflow with Claude Code, 0:34)

The organization has articulated a perspective on async agents — how Claude handles long-horizon, asynchronous tasks — positioning this as a distinct research and deployment concern. (Claude for Long-Horizon Tasks — Lance Martin, Anthropic, 0:31)

Pace of feature development

Anthropic's rapid release cadence is noted as a competitive challenge for third-party providers: the pace of new features is so high that providers building abstraction layers on top of Claude struggle to keep up. (To Give an Agent a Workspace, 1:13:19)