Three AI Tools Show Healthcare Moving Toward Domain-Specific Automation
Three new artificial intelligence products launched this month signal a shift in how the industry is deploying AI: away from general-purpose tools toward systems built for specific clinical and operational tasks.
Ambience Healthcare, Corti, and a partnership between Ensemble and Cohere each announced tools designed to solve narrow problems in healthcare workflows. The common thread is generative AI and large language models built or fine-tuned for healthcare rather than adapted from consumer-grade systems.
EHR-integrated nursing assistant launched
Ambience Healthcare expanded its Chart Chat tool to let nurses query Epic Systems electronic health records directly at the point of care. The feature retrieves patient medication histories, lab trends, and clinical information in seconds without forcing nurses to leave their EHR.
The tool operates inside the Ambience module within Epic and returns answers as text. Ambience said it is exploring integrations with other EHR systems beyond Epic.
Safety features include three-tier oversight: evaluations during deployment, real-time quality monitoring, and nurse feedback loops.
Medical coding model outperforms general AI systems
Corti released Symphony for Medical Coding, a system that analyzes clinical text and assigns diagnosis codes using a four-stage reasoning workflow. The system filters out irrelevant details, identifies active diagnoses, queries the ICD-10 index, and generates ranked code candidates with supporting evidence-mimicking how trained human coders work.
The underlying model, called Code Like Humans, was trained on 5.8 million electronic health records from 1.8 million patients. When tested across five datasets in U.S. and UK healthcare settings, it outperformed OpenAI and Anthropic models by more than 25%.
Corti said the advantage stems from treating medical coding as a reasoning problem rather than a labeling task. The system returns primary codes alongside alternatives and source text for audit purposes.
Symphony is available as an API, through Model Context Protocol, and via enterprise and cloud deployments.
Revenue cycle management gets custom model
Ensemble and Cohere announced they will build a custom language model tailored to healthcare revenue cycle management. The model will be fine-tuned on RCM tasks and embedded into AI agents that handle patient intake through account resolution.
Ensemble's operational expertise and process knowledge will shape how the model learns. The companies said it will not be trained on identifiable patient data or protected health information.
The partnership reflects a broader recognition that general-purpose AI systems miss regulatory nuances and healthcare-specific workflows. By pairing domain expertise with AI agents and automation, providers can reduce administrative friction and improve consistency.
Your membership also unlocks:
