AI Engineer (Healthcare)

Location: Remote, United States

Company

hirify.global provides specialty revenue cycle management solutions for healthcare organizations through its E360 RCM intelligent automation platform.

What you will do

• Build, train, fine-tune, and evaluate language models using supervised and reinforcement learning methods.
• Design reinforcement learning environments, reward functions, verifiers, experiments, and evaluation frameworks.
• Analyze source materials and datasets, create training and evaluation data, and apply automated and human validation.
• Run reproducible experiments, ablations, benchmarks, A/B tests, and model comparisons while diagnosing instability, reward hacking, and data leakage.
• Represent specialized domain knowledge in structured, testable formats and collaborate with healthcare domain experts.
• Document and present technical findings to engineers, domain experts, and business leaders.

Requirements

• Bachelor’s degree in computer science, machine learning, statistics, or a related field; equivalent hands-on experience is considered.
• 3+ years of experience training or fine-tuning language models.
• End-to-end reinforcement learning experience, including environment, reward function, and policy design.
• Strong Python and PyTorch skills, with experience in Hugging Face and parameter-efficient fine-tuning.
• Experience with LLM-based agents, multi-turn systems, language model evaluation, held-out datasets, and automated grading.
• Work must comply with HIPAA and hirify.global data policies, using de-identified or synthetic data wherever possible.

Nice to have

• Healthcare or highly regulated industry experience.
• Structured-output model training, advanced reward design, and a broad range of reinforcement learning algorithms.
• Knowledge graphs, ontologies, synthetic data generation and validation, or publications and open-source contributions.
• Experience handling protected health information under HIPAA.

Culture & Benefits

• Hands-on individual contributor role with ownership from experiment design through code, training, and results.
• Approximately half of the work focuses on reinforcement learning and model training, with the remainder covering knowledge engineering, data analysis, and evaluation.
• Close collaboration with domain experts and adaptation of data, training, and evaluation approaches as knowledge evolves.
• Commitment to hirify.global’s core values, vision, and mission.