New York, New York, United States
Hybrid
New York City or San Francisco | Hybrid Onsite | Full Time
Let's make high-quality healthcare free and accessible to everyone through AI. Doctronic's AI doctor already handles millions of consultations; our goal is to scale to billions while continuously improving clinical safety and accuracy. That means advancing how AI systems reason, learn, retrieve evidence, and earn trust in real clinical care.
Doctronic runs a real clinical practice, with patients consulting our AI doctor every day. That gives us a dataset no one else has to test ideas against. We're committed to publishing and open sourcing as we go.
You'll build the reasoning systems, learning methods, retrieval algorithms, and evaluation infrastructure that allow every component of Doctronic's clinical AI to improve with each iteration and earn greater autonomy over time.
This role blends research and engineering. What matters is real experimental or modeling work you can also ship, whether you come from applied ML or data science with strong engineering skills, or engineering with a research bent.
Design and implement the next generation of the architecture behind Doctronic's AI doctor, including reasoning, reflection, verification, tool use, routing, uncertainty handling, and escalation.
Build systems in which specialized agents and models work together to make clinical decisions safely, reliably, and efficiently, with an architecture that supports self-improvement over time.
Build the evaluation platform, rubrics, simulations, and experiments that measure how the AI doctor performs and know when a benchmark score rewards the wrong behavior.
Identify what should be improved, whether reasoning, retrieval, model behavior, data, or engineering, and determine whether each intervention genuinely made the system better.
Apply methods such as fine-tuning, distillation, reinforcement learning, preference optimization, and prompt and system optimization to improve specific components of the AI doctor.
Build the training data, feedback, reward, and experimentation pipelines that turn evaluation results and clinical expertise into system improvements.
Build search, ranking, retrieval, and grounding algorithms that connect clinical reasoning to trusted medical evidence, partner-specific content, and patient context.
Improve and evaluate retrieval based on its effect on downstream clinical decisions, not only on document-relevance metrics.
Wherever you're working, you'll own the problem end-to-end, from framing it to measuring whether it worked.
Builder-first org: senior engineers who take ownership of a problem and run with it
Autonomy here means moving fast on your own judgment and bringing in clinical, product, or engineering partners early when the problem calls for it, not after.
You are a research-minded engineer who wants to build intelligent systems and is equally serious about understanding and demonstrating their effectiveness.
You have deep experience in at least two of the following:
Agentic architectures, reasoning systems, and tool use
Model evaluation, experimentation, and rubric design
Model training, fine-tuning, distillation, or reinforcement learning
Search, ranking, retrieval, or RAG
You have strong ML fundamentals, including training data, objectives, metrics, failure analysis, calibration, and validation.
You're comfortable working in domains where ground truth is incomplete, and experts disagree, and where the right move is sometimes to question what "correct" even means before answering.
You think through the business impact of your work and scope and prioritize solutions accordingly.
You have strong engineering fundamentals, you build real systems, not just notebooks or one-off experiments.
You communicate clearly and collaborate well across engineering, clinical, and product teams.
Typically, candidates will have an advanced degree in a quantitative, computational, or scientific discipline and 3+ years of highly relevant applied or research experience in AI/ML; or 7+ years of relevant experience building and researching ML systems, including recent hands-on work with LLMs and generative AI.
We care more about the depth of your work than a specific credential or career path.
Published research, patents, meaningful open-source contributions, or novel production ML systems
Experience building AI systems at an early-stage or high-growth company
Experience in healthcare, clinical AI, or another regulated or safety-critical domain, including familiarity with clinical workflows, healthcare data, or standards such as HIPAA, FHIR, EHR, and HL7
Experience with human-feedback systems, RLHF, simulation, or synthetic-data generation
Work in AI safety, bias detection, calibration, or fairness
Competitive salary plus meaningful equity with real upside as we grow
Build AI systems transforming healthcare for millions
High autonomy and end-to-end ownership on problems that matter
Work side by side with physician-scientists
Health benefits
Parental Leave: 12 weeks fully paid parental leave for all parents, regardless of gender or path to parenthood — no distinction between birthing and non-birthing parent
Doctronic is backed by Union Square Ventures, Lightspeed Venture Partners, and Abstract Ventures, with three rounds of financing completed between February 2025 and January 2026.
Doctronic is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees.
Want to know if this job is worth applying to?
