TLDR
Hello Heart was an early adopter of AI in digital cardiovascular health, introducing its first AI-enabled feature in 2022 and expanding its use of the technology over time. Today, AI supports parts of the Hello Heart member experience, including our heart health AI chatbot, Nia, as well as predictive risk flags.
An early step into AI-powered heart health
Hello Heart was founded in 2013 to help people understand and manage their cardiovascular health. The platform brought blood pressure readings and other heart health information into an app where members could track their numbers, receive personalized coaching, and better understand how daily habits might affect their health.
AI was not the basis of the company at its founding. It became part of the platform as the technology matured and Hello Heart identified practical ways to use it in support of members.
In November 2022, Hello Heart announced Dot-to-Dot, a feature that uses AI to show members personalized correlations between behaviors—such as walking or taking medication—and their blood pressure. The goal was straightforward: help people see how their everyday choices may relate to changes in their heart health.
That launch marked a documented early use of AI within the Hello Heart platform. It also reflected an approach the company continues to follow: applying AI to defined cardiovascular use cases rather than treating the technology as an end in itself.
Building on years of cardiovascular data and experience
Hello Heart’s ability to explore new uses of AI is supported by something the company had been developing long before 2022: experience helping members track and manage cardiovascular risk factors.
Through connected blood pressure monitors and the Hello Heart app, members can record blood pressure and heart rate measurements and track information such as medications, activity, weight, and cholesterol. With appropriate privacy and security protections, these types of data can support research into how digital tools and analytical models may help people better understand cardiovascular risk.
From personalized insights to cardiovascular risk research
Hello Heart’s work with AI has expanded beyond showing members connections between habits and blood pressure.
In 2024, researchers from Hello Heart, Sheba Medical Center, and Scripps Research Translational Institute published a peer-reviewed study of a machine-learning model for short-term cardiovascular risk prediction. The retrospective study included 51,127 Hello Heart participants with hypertension and used electronic health record data along with measurements collected through at-home blood pressure monitors.
In the study population, the model showed stronger discrimination of 90-day and 365-day risk for heart attack and stroke than the traditional risk-assessment tools against which it was compared.
That qualification matters. A promising research result is not the same as a universal guarantee, and responsible use of AI in healthcare requires continued evaluation. The study nevertheless demonstrates how Hello Heart has applied machine learning to a defined cardiovascular question and evaluated the approach through peer-reviewed research.
Expanding AI-supported member guidance
Hello Heart continued that evolution with the 2025 introduction of Nia, its AI heart health assistant. Nia is integrated into the Hello Heart app and is designed to provide members with heart health and medication information, including personalized insights based on their experience in the program.
Nia adds a conversational way for members to access support alongside connected monitoring, app-based tracking and coaching, and medication adherence tools.
Within Hello Heart, different technologies and forms of support serve different purposes. Connected devices help capture measurements. The app helps members track and understand their information. AI can help identify patterns, flag risks, and make guidance easier to access. Clinical experts contribute to the evidence, rules, and review processes surrounding the experience.
What responsible AI adoption looks like
For employers and health plans, the most useful question is not whether a digital health platform has AI capabilities. It is how the platform uses AI, what evidence supports those uses, and what safeguards are in place.
When evaluating an AI-powered digital health platform, benefits leaders can ask:
- What problem is the AI intended to solve? The use case should be clearly defined and connected to a real member need.
- What data informs the technology? Vendors should explain what types of data are used and why those data are relevant to the intended purpose.
- How has the technology been evaluated? Look for evidence tied to the specific product or model, including peer-reviewed research where appropriate.
- What role do clinical experts play? Vendors should be able to explain how clinical expertise informs development, oversight, and escalation.
- What are the limits of the AI? Clear boundaries are especially important when technology is used to provide health-related information.
- How is member data protected? Ask how information is stored, used, shared, and protected, including whether it is used to train outside models.
These questions reveal more than a label can. They help benefits leaders distinguish between a clear, evidence-informed application of AI and a broad marketing claim.
FAQs
When did Hello Heart begin using AI?
Hello Heart publicly announced Dot-to-Dot, an AI-enabled feature designed to show personalized relationships between behaviors and blood pressure, in November 2022. The company later published research on machine-learning-based cardiovascular risk prediction and introduced Nia, its AI heart health assistant.
How does Hello Heart use AI today?
Hello Heart uses AI in defined areas of its cardiovascular platform, including personalized insights, cardiovascular risk flags, and conversational heart health and medication support. These capabilities work alongside connected monitoring, tracking and coaching tools, and clinical expertise.
Has Hello Heart’s AI been studied?
A 2024 peer-reviewed study evaluated a Hello Heart machine-learning model designed to predict short-term risk of heart attack and stroke among 51,127 participants with hypertension. In that study population, the model showed stronger discrimination of 90-day and 365-day risk for heart attack and stroke than the traditional risk-assessment tools against which it was compared.
What should employers ask about AI-powered digital health platforms?
Employers should ask what the AI is designed to do, what data inform it, how it has been evaluated, how clinicians contribute to its development and oversight, what its limitations are, and how member information is protected.
