AI Engineer, dexter

• We care less about formal seniority and more about speed, intelligence, ownership, and product impact. We like people who operate with founder-like urgency: they figure things out, make progress under ambiguity, and care about outcomes more than titles
• Strong software engineering skills, especially in Python or a comparable backend language
• Experience building AI-powered product features or LLM-based workflows
• Ability to turn ambiguous product ideas into working software quickly
• Strong understanding of how to design, test, validate, and improve AI outputs
• Experience integrating AI workflows into production systems
• Strong debugging instincts across application logic, data, model outputs, and user-facing behavior
• Experience with evaluation, testing, validation, or quality checks for AI-generated outputs
• Serious experience using AI development tools such as Claude Code, Codex, Cursor, Copilot, or similar as part of your daily workflow
• High intelligence, fast learning speed, bias to action, and ownership mindset
• Comfortable with ambiguity, fast decisions, and a high-trust startup environment
• Clear written and spoken English

Desirable

• At dexter health, we build AI-powered software for care teams. Our mission is to reduce administrative workload in healthcare so caregivers can spend more time with patients
• We are looking for a high-agency AI Engineer to help us build new AI features faster and improve the quality, reliability, and speed of existing AI workflows
• This is a hands-on engineering role. Not research. Not prompt-only. You will turn ambiguous product ideas into working, production-ready AI features
• The role is mostly focused on shipping AI product features, with some AI systems and infrastructure work where needed
• You will work closely with product and engineering, but we expect you to take real ownership. You should be comfortable moving fast, making technical decisions under ambiguity, and using AI development tools such as Claude Code, Codex, Cursor, Copilot, or similar every day
• Build new AI-powered product features from idea to production
• Improve existing AI workflows for quality, reliability, latency, and user value
• Design and implement LLM-based workflows, structured outputs, validation logic, and fallback behavior
• Build evaluation loops, tests, and quality checks for AI-generated outputs
• Integrate AI capabilities into existing product and backend systems
• Work with commercial and open-source LLMs without being tied to one specific provider
• Support self-hosted model workflows where they make sense for quality, speed, cost, or control
• Debug AI feature failures across inputs, outputs, data, backend logic, and user flows
• Use AI development tools as a core part of your daily workflow
• Ship quickly while keeping production quality high