Senior Machine Learning Engineer at One Medical
Colin Waldman is a software engineer on the Machine Learning team at One Medical, primarily focused on designing systems that simplify NLP model experimentation and productionization, and building predictive models that improve the patient-provider relationship. Prior to One Medical, Colin worked as an engineer and data scientist across multiple industries including fintech, healthcare and automotive. Colin received his BS in Mechanical Engineering, and his MS in Automotive Engineering, from The Ohio State University Center for Automotive Research.
WATCH LIVE: February 17 @ 12:40PM – 1:10PM ET
Electronic health records contain a wealth of information, and are frequently the primary tool that clinicians use when diagnosing and treating patients. Much of this information, however, is stored as unstructured text, forcing the clinician to comb through significant quantities of office notes, lab reports and other documents in order to find pertinent information.
This presents an interesting opportunity for natural language processing and machine learning – if labels can be assigned to the “pertinent” category (ex. indications of dementia) in an efficient and reliable manner, supervised ML models can be developed which automatically extract this information and surface it directly to the clinician. These models result in less screen time, and more face-to-face time with the patient.
This presentation outlines how One Medical’s Machine Learning team has developed an end-to-end pipeline capable of labeling and validating data, running machine learning experiments, and hosting predictive models, all with only two configuration files and a minimal amount of code.