AI Harness Engineer (100% Remote, Worldwide)

Tether is a financial technology company at the forefront of blockchain innovation and digital asset infrastructure. Operating across multiple domains—from stablecoins and energy solutions to AI-powered applications and education—Tether serves hundreds of millions of users globally. The company is building the next generation of local AI assistants, and this role is central to that mission.

This is an opportunity to join Tether's Engineering team as an AI Harness Engineer and help scale cutting-edge peer-to-peer technology for artificial intelligence. You will work on QVAC, Tether's platform for connecting users to local AI, designing systems that enable AI to operate reliably across diverse hardware and network conditions. Your contributions will directly impact how millions of people interact with AI in their daily lives.

What you'll do

• Build and ship new features and bug fixes with emphasis on system reliability, performance, and user experience
• Design new tools and modules that expand AI capabilities while supporting different model types and sizes
• Establish and maintain high code quality standards through code reviews, consistent practices, and automated testing
• Develop peer-to-peer technology infrastructure for local AI deployment and operation
• Create evaluation frameworks and test systems for non-deterministic AI systems
• Design systems that handle failure gracefully, including network disruptions, model errors, timeouts, and recovery logic

What you'll bring

• Five or more years of production software development, with substantial experience in peer-to-peer systems, AI and large language model tooling, or both
• Hands-on experience building AI agent harnesses, including tool calling, agent loops, context management, and structured output generation
• Strong proficiency in JavaScript and TypeScript, with demonstrated ability to model complex state across sync protocols and agent workflows
• Proven experience building evaluation and testing infrastructure for systems with non-deterministic behavior
• Background designing systems for failure scenarios involving network partitions, model errors, timeouts, and retry mechanisms

Nice to have

• Open-source contributions to peer-to-peer, networking, or AI tooling projects
• Familiarity with local model deployment and optimization

What they offer

• Full-time employment with a 100% remote position available to candidates worldwide
• Opportunity to work on infrastructure serving the rapidly expanding local AI assistant market
• Access to cutting-edge peer-to-peer and AI technology development