Elevate Patient Education with HeartPatientAI

HPAI is a HIPAA-compliant Generative AI Patient Education platform built on the latest AI technology. Supported by an extensive cardiovascular-specific video library and additional printable, custom-branded educational tools, HeartPatientAI is the cutting edge of online learning for heart patients.

Large Language Models (LLM), are a specific type of artificial intelligence software that uses deep learning algorithms designed to understand and generate human-like textual output based on the data it has been trained on. These models are called “large” due to the vast amount of data they are trained on, allowing them to comprehend and produce language across a wide range of topics and contexts. Like the human brain, large language models must be pre-trained and then fine-tuned so that they can solve text classification, question answering, document summarization, and text generation problems.

HPAI uses the latest publicly available LLM technology, architectures, and models. From that, we have created our ‘Custom Content’ LLM. HPAI’s Custom Content LLM is built on the cardiovascular and medical knowledge base. A Domain-specific LLM is a general model that is provided relevant contextual data and instructions to perform well-defined tasks required by industry or organizational guidelines. Domain-specific LLMs are designed to address the limitations of Generic LLMs in specialized fields.

HPAI’s Custom Content LLM is a specialized Domain Specific LLM that has been provided a set of curated data. This customization process enables the model to generate Output that is more closely aligned with the unique needs or preferences of a particular user or organization. For example, HPAI can provide our Custom Content LLM a hospital’s specific protocols and patient information to ensure the model can accurately answer questions or write content that reflects particular cardiovascular terminology and information.

HPAI uses a secure ‘private’ Generative AI workflow engine (VectorShift) to build the core generative AI answering engine. HPAI only use vetted cardiovascular and medical content provided in collaboration with the American College of Cardiology (ACC), the National Institutes of Health, the National Library of Medicine, medical device manufacturer product-specific patient guides, documentation, and pharmaceutical data as context for its foundational models.

It’s important to understand that HPAI only uses its ‘custom content LLM’ to power its generative AI Output. This gives patients and clinicians the confidence the output is accurate and does not include hallucinations.

HPAI Output is based only on provided clinically vetted content. Our clinician beta testing has shown it to be highly accurate.

HPAI’s AI workflow leverages multiple vector databases that identify the most relevant content based on a user query, using both semantic and lexical search techniques. We then used the latest generative AI techniques to do what’s known as ‘Retrieval Augmented Generation’ and ‘Prompt Engineering’ to ensure the Output is highly accurate and follows strict guidelines.

We take data security seriously and built our platform with security in mind. HPAI is HIPAA compliant. User data is end-to-end encrypted. We have “zero data retention” to ensure that data is never stored or used for training purposes. So when you close out the session, everything is deleted. HPAI uses the Vanta Security platform for SOC 2 Type 2, and HIPAA Compliance.