Clinical AI for prior authorization and payer-provider alignment
Cohere Health builds clinical intelligence systems focused on prior authorization—the administrative bottleneck between payers and providers. The stack reveals an ML-heavy organization: PyTorch, transformers, Cohere (their own model), SageMaker, plus a strong data infrastructure layer (Kafka, Iceberg, dbt, Snowflake, Redshift, Airflow). Active projects span production ML deployment, retrieval and generative algorithms, and a payment integrity suite, while hiring accelerates across engineering and data roles—indicating scaling toward real-time clinical decisioning at scale.
Notable leadership hires: Medical Director, Associate Medical Director, Program Director, Architecture Director
Cohere Health operates in healthcare administration, specifically automating prior authorization—the process by which physicians and health plans align on care decisions. Founded in 2019 and based in Boston, the company serves payers and providers by reducing administrative friction while improving patient outcomes. The product surface includes intelligent prior authorization decisioning, clinical findings extraction via ML, and a payment integrity suite. The organization spans 501–1,000 employees across engineering, data, healthcare (clinical), operations, and product functions, with notable leadership in medical direction and architecture—reflecting both clinical domain expertise and infrastructure maturity.
The stack includes PyTorch, transformers, and Cohere (their own model), deployed via SageMaker. Current projects focus on retrieval, classification, and generative use cases for clinical decision-making.
Cohere uses Kafka, Iceberg, dbt, Snowflake, Redshift, AWS EMR, AWS Glue, Athena, and Apache Airflow—a modern data lakehouse and ETL infrastructure typical of ML-heavy production organizations.
Other companies in the same industry, closest in size