Block builds products such as Square and Cash App, while its AI capabilities team partners across the organization to develop customer-facing features and internal tools like Buzz, an open-source workspace for human-agent teams. According to Bradley Axen, Head of AI Capabilities, frontier models like Claude Fable 5 have transformed software engineering at Block by enabling company-wide code migrations that span hundreds or thousands of pull requests across multiple repositories. Rather than applying top-tier frontier models to every task, Block uses model orchestration. Fable handles high-level design work such as data models, API specs, and algorithms, while directing up to dozens of smaller worker models like Opus or Sonnet to execute file edits and run tests. This strategy balances frontier intelligence with smaller model efficiency. Block runs these multi-agent workflows within Buzz to maintain visibility into team progress. To manage resources, the company is building auto-selectors to help technical employees choose the right model and effort level for specific tasks. Despite high levels of agent autonomy in writing code and managing tech debt, Block maintains rigorous safeguards. Merges to main and production deployments require passing security checks along with dual human approval, ensuring that artificial intelligence systems cannot unilaterally deploy changes.
How Block orchestrates Claude Fable across thousands of pull requests
Bradley Axen, Head of AI Capabilities at Block, explains how the company uses Claude Fable 5 and other frontier models to orchestrate company-wide code migrations across thousands of pull requests. Block relies on its open-source workspace, Buzz, to manage human-agent collaboration and model orchestration. The company also employs strict human oversight, requiring security checks and dual human approval for code deployments.
Why it matters
This approach demonstrates how enterprises can leverage model orchestration to scale software development efficiently while maintaining cost-effectiveness and strict human safety controls.
Topic dossiers Anthropic & Claude