Software Engineer: AI Harnesses & Tooling

Description: Build a reliable agent harness for LLMs that can be extended by users and other agents. Work on problems like accountable memory, multi-agent collaboration, context retrieval, routing work across models, and trust permissions. Requirements: Strong understanding of agent runtimes, tool-calling loops, and multi-step task failures. Experience with LLM APIs at the protocol level, including streaming and structured outputs. Ability to design systems that improve through use and adapt to user needs. Familiarity with systems languages (Rust, Go) and TypeScript is a plus. Benefits: Competitive salary range of $215k to $325k plus meaningful equity. Opportunity to shape the product and culture as part of an early team in a founding role.