Show HN: Drift, write LLM agents in English and transpile to async Python (github.com)

Drift, a new framework for building AI agents, has launched its intent-based language designed for creating agentic systems using English-shaped blocks. Users can write agents that are transpiled into asynchronous Python code, enabling them to leverage large language models (LLMs) like "gpt-5.4-nano." An example agent, InboxTriage, demonstrates how to triage emails with conditional responses and parallel processing, all while keeping costs low—running five emails in just 1.82 seconds for $0.0092. The framework simplifies the development process by allowing immediate testing with mock providers and supporting the integration of real API keys from major LLM providers. This platform is significant for the AI/ML community as it democratizes the creation of intelligent agents, making it accessible to a broader audience without requiring deep programming expertise. The built-in features—ranging from multi-provider routing to structured error handling and memory management—encourage experimentation with parallel processing and complex workflows. By allowing users to define custom verbs and support for external tools, Drift enables more personalized and efficient agent development, potentially accelerating the integration of AI into various applications and enhancing productivity.