A healthcare integration has been developed using Bridgegate (BG) to connect with the Epic EHR sandbox environment. The application allows users to input a patient’s MRN through a simple UI, triggering a FHIR API request authenticated via JWT as a backend application. Upon request, clinical data such as Patient, Encounter, Observation, Medication, and Allergy resources are retrieved in FHIR JSON format. This structured data is then processed using AI to generate a human-readable clinical summary, providing users—such as clinicians or care coordinators—with an instant, concise view of the patient’s health history without navigating raw FHIR data.

We orchestrated multiple pipelines for a customer across multiple protocols and data types like CCD, FHIR, proprietary JSONs. This data was sliced to create a data lake to homogenize the received data to train the LLM model. This helped them in predictive analysis and insights on the patient, vitals, drug level so that these insights can translate into decision, backed by the trend visible on the dashboards. These dashboards can be built in as per requirement and the insights can be tailored to the function in the organization which is consuming these data points.

Our customer gets data from multiple healthcare facilities via HL7 feeds, API calls etc. These messages are for patient enrollment, hence making them critical in our customer application to move the patient ahead in the clinical journey. We developed an ecosystem, in case any facility stops sending us data in the last 24 hours, we trigger an alert to the facility to ensure they are pre informed, instead of realizing in the patient journey later via support tickets or escalations. The alerts escalate internally in our customer’s hierarchy to ensure in case any issue exists, it can be handled. We have also automated the requirement of fixing the issues via our automation processes, be it the socket tunnel reaffirmation or the API endpoint being unavailable.