About Ultralight
At Ultralight, our mission is to empower providers to help their patients live their healthiest lives today. We're building the future of personalized medicine by giving AI superpowers to clinicians. Think Claude Code or Cursor, but for doctors.
We are built a sophisticated AI platform that powers functional, integrative, and longevity medical practices. Our platform combines large language models, healthcare-specific agents, and human expertise to empower practitioners to deliver high-quality preventive care that learns, reasons, and adapts to each patient's unique needs.
We are powering a community of forward-thinking healthcare professionals committed to transforming the industry. Our journey is fueled by a passion for better healthcare—for providers and patients alike. Together with our partners, we're not just imagining the future of healthcare. We're creating it.
Role Overview
As a Software Engineer acting as the Tech Lead for Agent Orchestration, you will apply strong software engineering principles to architect, manage, and massively parallelize the workload of our specialized AI agents, reporting directly to our Co-Founder and CTO. You will effectively treat the agents as a team of "employees," focusing on task distribution, resource allocation, and ensuring their collective, high-performance output to accelerate our internal AI development and testing cycles. This role is instrumental in driving the design, performance tuning, and reliability of complex agentic systems and pipelines that underpin our production platform.
What You’ll Be Working On
- Architect and implement the agent orchestration layer responsible for massively parallelizing tasks, efficiently distributing workloads, and managing a fleet of healthcare-specific agents. This is key to ensuring reliability, throughput, and performance at scale.
- Function as the "Tech Lead" for the agents, designing and implementing the control plane to reliably chain together agent actions, reasoning, and tool use to accelerate internal R&D, experimentation, and complex testing workflows.
- Establish and maintain rigorous evaluation (Evals) frameworks to measure the individual performance and collaborative success of the agent team, ensuring high-quality, reliable, and scalable agent development.
- Drive technical decisions around AI architecture, model selection, and deployment strategies.
- Collaborate with the team on core platform development, ensuring seamless integration of AI components into our full-stack clinical application.1
- Provide feedback and ideas to improve specific procedures, tech implementations, and day-to-day operations to the rest of the team.1
- Fostering an environment of continuous learning, improvement, and excellence.
Who We’re Looking For
- A Software Engineer with 5+ years of experience, some of which focused on applied machine learning, LLM applications, or agentic systems.
- Proven experience designing and shipping production LLM applications and complex AI agents.
- Deep expertise in developing reliable retrieval techniques (like RAG) and robust agentic architectures/harnesses.
- Proficiency in building and automating AI evaluation pipelines (Evals) to measure agent performance, hallucination rates, and overall response quality.