Senior Software Engineer, AI

Tono Health•Brooklyn, NY

11h•$180,000 - $275,000•Hybrid

About The Position

Tono is building software and operations to make specialty care faster to access and easier to deliver, starting with dermatology. The company operates its own national telehealth practice and an in-person clinic in DUMBO, Brooklyn, and aims to build an autonomous EHR that can help specialty clinics across the country reduce administrative work. Tono currently serves patients in 33 states. The Senior Software Engineer, AI will own AI agents that support Tono’s clinical and operational workflows, including drafting clinical notes, triaging patient intake, coordinating care, and automating administrative work. The role also builds shared agent infrastructure, tool and memory abstractions, durable workflows, tracing, and evaluation systems. The engineer works in a TypeScript monorepo and collaborates closely with engineers, physicians, and care coordinators. The team ships iteratively with clear ownership, small pull requests, careful reviews, feature flags, safe migrations, and built-in observability. Engineers may work across frontend, backend, data, and infrastructure, with an emphasis on maintainable, reliable production software in a regulated environment. The position is hybrid in New York City, with a few days per week in the DUMBO, Brooklyn office; fully remote candidates may be considered.

Requirements

At least 5 years of software engineering experience.
Strong TypeScript experience across frontend and backend.
Hands-on experience building and shipping LLM-powered agents in production, rather than only prototypes.
Experience designing evaluations, including test sets, LLM-as-judge, regression suites, and connecting quality metrics to real outcomes.
Practical experience making LLMs reliable, including structured outputs, retries and fallbacks, tool-calling design, and monitoring.
Experience with data-labeling techniques and feedback loops that improve agents over time.
Solid relational database fundamentals, including PostgreSQL, schema design, and migrations.
Comfort shipping iteratively with feature flags, instrumentation, and tight feedback loops.
Ability to take an ambiguous problem, break it down, and drive it to completion.
Care about code quality and take pride in engineering craft.

Nice To Haves

Experience with agent frameworks such as Mastra or LangGraph.
Experience with durable workflow tools such as Inngest.
Experience with LoRA or fine-tuning.
Experience at startups, in healthcare or regulated industries, or with multi-tenant systems.

Responsibilities

Design, build, and ship AI agents that perform clinical and operational work, such as drafting clinical notes, triaging intake, coordinating care, and automating administrative tasks.
Build shared agent infrastructure, tool and memory abstractions, durable workflows, tracing, and evaluation harnesses for other engineers.
Create evaluations and labeled datasets using clinician feedback, and use them to identify regressions before release.
Build improvement loops that incorporate physician corrections into prompts, retrieval, and fine-tuning.
Add guardrails, human-in-the-loop review, and safe fallbacks so agents fail gracefully.
Debug unexpected agent behavior using traces, logs, and replay.
Build reliable production AI systems using structured outputs, evaluation loops, monitoring, and escalation to humans when appropriate.
Work across frontend, backend, data, and infrastructure as needed.
Ship iteratively using feature flags, instrumentation, careful reviews, and safe migrations.
Maintain and improve software quality, including dependency upgrades, performance work, and removal of dead code.
Use logs, metrics, alerting, and error monitoring as part of the engineering workflow.