About the Role

Synthio is looking for exceptional AI Engineers who can build deeply technical systems and communicate clearly about what they are building. You will work across our AI stack, from LLM orchestration and retrieval to agents, evaluations, voice systems, APIs, and production infrastructure, owning problems from exploration through deployment and iteration.

What You’ll Do

• Build production AI systems using frontier models, retrieval, tool use, agents, and structured workflows.
• Design and productionize LLM pipelines, including prompting, context engineering, retrieval, orchestration, and evaluations.
• Build backend services, APIs, data flows, and integrations that power AI products.
• Work on real-time voice and multimodal AI systems.
• Design evaluation suites that identify failures, regressions, and opportunities to improve model behavior.
• Experiment with new models and techniques and determine what is useful in production.
• Debug problems across models, application code, data, and infrastructure.
• Communicate technical decisions, tradeoffs, results, and failures clearly

What We’re Looking For

• 3+ years of strong software or AI engineering experience.
• Excellent Python and software engineering fundamentals.
• Experience shipping LLM-powered or machine-learning systems into production.
• Strong understanding of retrieval, embeddings, tool calling, agents, and evaluations.
• Ability to reason about latency, reliability, observability, cost, and failure modes.
• Exceptional written and verbal communication.
• Ability to take an ambiguous problem, structure it clearly, and drive it to completion.
• High agency and a bias toward building and shipping.

Strong Plus

• Experience with speech, voice AI, or real-time systems.
• Experience with AWS and production cloud infrastructure.
• Experience building developer platforms or reusable AI infrastructure.
• Experience working directly with enterprise customers or in regulated industries

Why This Role Matters

At Synthio, we build software that directly supports patients, physicians, and pharma teams. This is production software, not prototypes or demos. The systems you build will be used in real workflows, under real constraints, and at real scale. You’ll have meaningful ownership and a clear line of sight from your code to real-world impact.