Job description

ABOUT US Epsilon Health is transforming medical imaging and diagnostics by bringing frontier AI into clinical practice. We're tackling some of healthcare's most critical challenges, building technology that directly improves patient outcomes. This year, we've served over 300,000 patients and now serve 3,000+ daily, working with some of the largest medical institutions in the U.S. By year-end, we're on track to interpret 1% of all daily U.S. X-rays. Our research team brings experience from DeepMind and Meta, powered by one of the industry's most comprehensive proprietary medical imaging datasets, continuously enriched by real-world clinical data. Together, we're building a future where frontier AI helps save millions of lives today, not tomorrow. Recently out of stealth, we're proudly backed by Jack Altman, Uncork Capital, Renegade Partners, and AlleyCorp, one of NYC's most active healthcare venture firms. ABOUT THE ROLE We're looking for an exceptional Research Engineer or Research Scientist to advance how we curate data and evaluate frontier AI models for medical imaging. You'll own the systems and methodologies that determine the quality of our training data and how we measure model performance, building the foundation for reliable, clinically accurate AI. You'll tackle fundamental challenges in data-centric machine learning, multimodal evaluation, and clinical reliability, from identifying subtle dataset errors to developing rigorous benchmarks for radiology report generation. Unlike traditional research labs, you'll work with continuously growing real-world clinical data and feedback from practicing radiologists, developing evaluation methods that connect offline research performance with real-world clinical outcomes. You'll work across our research team, collaborating closely with scientists and engineers developing vision foundation models, VLMs, and post-training methods. This is a highly autonomous role with the opportunity to identify critical research bottlenecks and solve them end to end. WHAT YOU'LL DO - Build the foundation for better models. Develop scalable data filtering, curation, and quality pipelines that improve the reliability of training datasets across medical imaging modalities. - Advance data-centric AI research. Explore novel approaches to detecting label noise, image-report misalignment, duplicates, and other subtle dataset quality issues. - Redefine how AI is evaluated. Design rigorous evaluation frameworks for radiology report generation, measuring clinical accuracy, completeness, hallucinations, and factual consistency. - Connect AI evaluation with clinical expertise. Work with radiologists to translate clinical judgment and expert feedback into scalable data quality and evaluation signals. - Develop intelligent evaluation systems. Build model-based evaluators, automated grading methods, and human-in-the-loop workflows that improve how we measure model quality. - Own research infrastructure end to end. Develop benchmarks, evaluation pipelines, and continuous regression testing that help researchers iterate quickly and confidently. - Drive research across the organization. Identify bottlenecks across data, training, and evaluation, developing tools and methodologies that accelerate the entire research team. - Contribute to the research community. Share technical insights, develop best practices, and contribute to publications or open-source research. WHAT WE'RE LOOKING FOR We care more about what you've built and discovered than where you've worked or how many years you've spent doing it. - Strong expertise in machine learning, data-centric AI, model evaluation, or research engineering. - Demonstrated ability to design and build robust data pipelines, evaluation systems, or ML research infrastructure. - Strong Python programming skills and the ability to develop reliable, scalable research tooling. - Experience in at least one of our core areas: dataset curation and quality, or evaluation of generative and multimodal models. - Ability to independently identify research bottlenecks, formulate solutions, and execute projects end to end. - Strong engineering fundamentals and experience working with complex datasets or machine learning systems. - A rigorous, experimental mindset and the ambition to solve difficult, unsolved problems. NICE TO HAVE - Experience developing evaluation frameworks for LLMs, VLMs, or other generative models. - Familiarity with data filtering, deduplication, label-noise detection, or active learning. - Experience building model-based evaluators, LLM-as-judge systems, or human-in-the-loop annotation workflows. - Knowledge of clinical evaluation methodologies, including factuality assessment, hallucination detection, or RadGraph-style metrics. - Background in medical imaging, radiology, clinical NLP, or other healthcare AI applications. - Experience working with large-scale data processing tools such as Spark, Airflow, or Databricks. - Publications or open-source contributions in machine learning, data-centric AI, evaluation, or medical AI. We welcome exceptional researchers and engineers from both academic and industry backgrounds. A traditional career path or prior healthcare experience is not required.

Requirements

Must-haves

We're looking for an exceptional Research Engineer or Research Scientist to advance how we curate data and evaluate frontier AI models for medical imaging.
You'll own the systems and methodologies that determine the quality of our training data and how we measure model performance, building the foundation for reliable, clinically accurate AI.
Build the foundation for better models. Develop scalable data filtering, curation, and quality pipelines that improve the reliability of training datasets across medical imaging modalities.
Advance data-centric AI research. Explore novel approaches to detecting label noise, image-report misalignment, duplicates, and other subtle dataset quality issues.
Redefine how AI is evaluated. Design rigorous evaluation frameworks for radiology report generation, measuring clinical accuracy, completeness, hallucinations, and factual consistency.
Connect AI evaluation with clinical expertise. Work with radiologists to translate clinical judgment and expert feedback into scalable data quality and evaluation signals.
Develop intelligent evaluation systems. Build model-based evaluators, automated grading methods, and human-in-the-loop workflows that improve how we measure model quality.
Own research infrastructure end to end. Develop benchmarks, evaluation pipelines, and continuous regression testing that help researchers iterate quickly and confidently.
Drive research across the organization. Identify bottlenecks across data, training, and evaluation, developing tools and methodologies that accelerate the entire research team.
Contribute to the research community. Share technical insights, develop best practices, and contribute to publications or open-source research.
We care more about what you've built and discovered than where you've worked or how many years you've spent doing it.
Strong expertise in machine learning, data-centric AI, model evaluation, or research engineering.
Demonstrated ability to design and build robust data pipelines, evaluation systems, or ML research infrastructure.
Strong Python programming skills and the ability to develop reliable, scalable research tooling.
Experience in at least one of our core areas: dataset curation and quality, or evaluation of generative and multimodal models.
Ability to independently identify research bottlenecks, formulate solutions, and execute projects end to end.
Strong engineering fundamentals and experience working with complex datasets or machine learning systems.

Nice-to-haves

Experience developing evaluation frameworks for LLMs, VLMs, or other generative models.
Familiarity with data filtering, deduplication, label-noise detection, or active learning.
Experience building model-based evaluators, LLM-as-judge systems, or human-in-the-loop annotation workflows.
Knowledge of clinical evaluation methodologies, including factuality assessment, hallucination detection, or RadGraph-style metrics.
Background in medical imaging, radiology, clinical NLP, or other healthcare AI applications.
Experience working with large-scale data processing tools such as Spark, Airflow, or Databricks.
Publications or open-source contributions in machine learning, data-centric AI, evaluation, or medical AI.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aimachine learningmlpythonllm

Platform: ashby
Effort: quick

Benefits

Benefits detected in the listing.

Vision