Project Mulberry: Apple’s secret AI-powered health coach

Lloyd Price
Apr 16
9 min read

Exec Summary

Project Mulberry is Apple’s codename for an ambitious AI-driven healthcare initiative aimed at transforming its Health app into a comprehensive, personalized health platform. The project, expected to debut with iOS 19.4 in spring or summer 2026, integrates an AI-powered health coach designed to function like a virtual doctor, offering tailored recommendations based on data collected from Apple devices such as the iPhone, Apple Watch, and earbuds.

Here’s a detailed overview based on available information:

Key Features of Project Mulberry

AI Health Coach

The core of Mulberry is an AI agent trained to replicate aspects of a real doctor’s guidance. It analyzes health metrics like heart rate, sleep patterns, activity levels, and blood oxygen to provide personalized advice on improving wellness.

The AI is being developed with input from Apple’s in-house physicians and will incorporate data from external medical experts in fields like sleep, nutrition, physical therapy, mental health, and cardiology.

Revamped Health App

The Health app will undergo a major overhaul, shifting from a passive data repository to an active health management tool. It will include a redesigned interface and new features like advanced food tracking to monitor dietary habits.

The app may integrate with other Apple services, such as Fitness+, and use device cameras to analyze workout forms and provide real-time feedback.

Educational Content

Apple is collaborating with medical experts to produce video content explaining health trends and conditions. These videos will be filmed at a new facility near Oakland, California, and may feature a prominent doctor as the “face” of the service, tentatively called “Health+.”

Potential Insurance Integration

There are indications that Apple may partner with national lab chains like Quest or Labcorp to allow the AI coach to order lab tests, with costs potentially billed to users’ health insurance, depending on coverage.

Development and Leadership

Evolved from an earlier initiative called Project Quartz, Mulberry has become a top priority for Apple’s health division, led by Dr. Sumbul Desai, Apple’s VP of Health, with significant involvement from COO Jeff Williams.

The AI is being trained with real-world medical data from Apple’s team of physicians, ensuring credibility and accuracy.

Development is in full swing, with a planned reveal possibly at WWDC on June 9, 2025, and a public release in 2026.

Strategic Context

Apple’s Health Vision: CEO Tim Cook has repeatedly emphasized that healthcare is Apple’s greatest societal contribution. Mulberry aligns with this by leveraging Apple’s ecosystem, particularly the Apple Watch’s health sensors, to shift from reactive data collection to proactive health management.

Market Positioning: The global AI healthcare market is projected to grow significantly, potentially reaching $150–670 billion by 2030–2034. Mulberry positions Apple to capture a share of this market while enhancing its ecosystem’s value.

Competition: Similar AI health coaching tools exist (e.g., Google’s Fitbit Personal AI, ExerRing), but Apple’s integration of hardware, software, and medical expertise could set it apart.

Challenges and Concerns

Accuracy and Reliability: Ensuring the AI’s recommendations are medically sound is critical, as errors could undermine trust or lead to health risks.

Privacy: Handling sensitive health data raises privacy concerns, though Apple emphasizes user consent and secure data practices.

User Adoption: Some skepticism exists about whether users want AI-driven health advice, with critics questioning if it’s being “pushed” without demand.

Medical Role: It’s unclear if the AI will focus on wellness coaching or attempt diagnostic capabilities, which could risk users delaying professional care.

Public Sentiment

Some users are excited about AI improving healthcare access, especially given frustrations with traditional systems, while others doubt its practicality or fear it’s overhyped “vaporware.”

Enthusiasm centres on personalised coaching and potential cost savings, but concerns include over reliance on wearable data and whether the AI can truly replace human doctors.

Broader Implications

Project Mulberry could redefine how users interact with health technology, making preventive care more accessible and integrated into daily life. If successful, it may set new standards for privacy and reliability in AI healthcare while reinforcing Apple’s role as a wellness platform, not just a tech company.

However, its impact depends on overcoming technical, ethical, and regulatory hurdles.

Nelson Advisors > HealthTech M&A

Nelson Advisors specialise in mergers, acquisitions and partnerships for Digital Health, HealthTech, Health IT, Healthcare Cybersecurity, Healthcare AI companies based in the UK, Europe and North America. www.nelsonadvisors.co.uk

We work with our clients to assess whether they should 'Build, Buy, Partner or Sell' in order to maximise shareholder value and investment returns. Email lloyd@nelsonadvisors.co.uk

Nelson Advisors regularly publish Healthcare Technology thought leadership articles covering market insights, trends, analysis & predictions @ https://www.healthcare.digital

We share our views on the latest Healthcare Technology mergers, acquisitions and partnerships with insights, analysis and predictions in our LinkedIn Newsletter every week, subscribe today! https://lnkd.in/e5hTp_xb

#HealthTech #DigitalHealth #HealthIT #NelsonAdvisors #Mergers #Acquisitions #Growth #Strategy #Cybersecurity #HealthcareAI #Partnerships #NHS #UK #Europe #USA #Canada

Nelson Advisors

Hale House, 76-78 Portland Place, Marylebone, London, W1B 1NT

Contact Us

lloyd@nelsonadvisors.co.uk

Meet Us

Digital Health Rewired > 18-19th March 2025

NHS ConfedExpo > 11-12th June 2025

HLTH Europe > 16-19th June 2025

HIMSS AI in Healthcare > 10-11th July 2025

Project Mulberry potential timelines

Apple's Project Mulberry, aimed at integrating an AI-powered health coach into the Health app, has the following potential timeline based on available information:

Launch Target: Expected with iOS 19.4, slated for spring or summer 2026.
Development Phase: Currently in progress, with Apple training the AI using data from hired physicians and external health experts.
Announcement: Likely to be revealed at WWDC 2025 (typically June), aligning with iOS 19 previews, though specific dates are unconfirmed.

These timelines are based on reports from sources like Bloomberg and may shift due to development or regulatory factors.

Potential Features of Project Mulberry

AI-Powered Health Coach: An AI agent integrated into the Health app, trained on data from Apple-hired physicians and experts in nutrition, cardiology, and other fields, to provide personalised health guidance.
Health Data Analysis: Leverages data from iPhone, Apple Watch, and other wearables to offer tailored recommendations on diet, exercise, sleep, and wellness.
Food Tracking: Allows users to log meals for nutritional insights and dietary advice.
Workout Form Analysis: Uses iPhone cameras to evaluate exercise form and provide real-time feedback.
Educational Content: Includes videos and resources from doctors, potentially hosted by a prominent physician for a “Health+” service.
Proactive Wellness Plans: Generates customised health plans, such as fitness routines or stress management strategies, based on user data.
Integration with Apple Ecosystem: Seamlessly works with existing health features like heart rate monitoring, sleep tracking, and activity rings.

Potential Benefits of Project Mulberry

Personalised Health Guidance: Offers users tailored advice, making health management more accessible and proactive without needing constant doctor visits.
Improved Health Outcomes: Encourages better lifestyle choices through data-driven insights, potentially reducing risks of chronic conditions.
Convenience: Centralises health tracking and coaching within the Apple ecosystem, streamlining user experience.
Educational Empowerment: Provides expert-backed content to help users understand and manage their health better.
Scalability: Could reach millions of Apple device users, democratising access to health coaching.
Motivational Support: AI-driven feedback and plans may boost user engagement in fitness and wellness goals.

While promising, the project faces challenges like ensuring AI accuracy, protecting user privacy, and navigating regulatory requirements for health-related advice. It aims to complement, not replace, professional medical care.

Technology stack behind Apple's Project Mulberry

Specific details about the technology stack behind Apple's Project Mulberry are not publicly disclosed, as Apple maintains strict confidentiality around its in-development projects. However, based on available reports and Apple’s existing technological ecosystem, we can make informed inferences about the likely components of the technology stack for Project Mulberry, which involves a revamped Health app with an AI-powered health coach.

Below is an analysis of the potential technologies involved:

1. Artificial Intelligence and Machine Learning

AI Agent / Large Language Model (LLM): Project Mulberry’s core feature is an AI-powered health coach that analyses user health data and provides personalized recommendations. This likely leverages Apple’s Apple Intelligence framework, introduced with iOS 18, which includes generative AI capabilities. The AI agent is reportedly trained on data from Apple-hired physicians and external experts, suggesting a custom-trained model tailored for health applications.
On-Device Machine Learning: Apple emphasizes privacy, so much of the AI processing is expected to occur on-device using Core ML, Apple’s machine learning framework. This allows the Health app to analyse sensitive health data locally, minimising cloud dependency.
Health Data Analysis: The AI likely uses advanced algorithms to interpret metrics like heart rate, sleep patterns, and activity levels, possibly employing natural language processing (NLP) for user interactions and predictive analytics for health trend detection.

2. Health App and iOS Integration

iOS 19.4: Project Mulberry is slated to launch with iOS 19.4, indicating the Health app will be built using Swift and SwiftUI, Apple’s primary programming language and UI framework for iOS apps. These ensure a seamless, native user experience across iPhone, Apple Watch, and other devices.
HealthKit: Apple’s HealthKit framework is central to collecting and managing health data from devices like the Apple Watch, iPhone, and third-party accessories. It securely stores and shares data (with user consent) between the Health app and the AI coach.
Camera Integration: The app’s workout form analysis feature uses the iPhone’s rear cameras, likely leveraging Vision Framework for computer vision and ARKit for motion tracking to provide real-time feedback on exercise technique.

3. Hardware and Sensors

Apple Watch and Wearables: The AI coach relies on data from devices like the Apple Watch, which uses sensors for heart rate, blood oxygen, ECG, and motion tracking. These are powered by Apple’s Sensor Fusion technology and CoreMotion APIs to aggregate real-time biometric data.
iPhone Cameras: For workout analysis, the iPhone’s camera system, combined with LiDAR (on Pro models) and neural processing via the A-series chip or M-series chip (for future iPads), enables precise movement tracking.
Powerbeats Pro 2 and AirPods: Emerging health features, like heart rate sensors in Powerbeats Pro 2, suggest integration with Bluetooth Low Energy (BLE) and Apple’s H1/H2 chips for seamless data transfer.

4. Data Privacy and Security

Secure Enclave and Differential Privacy: Apple’s commitment to user privacy means health data is likely encrypted and stored in the Secure Enclave on Apple devices. Differential Privacy techniques may be used to anonymise data for AI training while maintaining accuracy.
On-Device Processing: To protect sensitive health information, most AI computations are expected to run locally, reducing reliance on cloud servers and aligning with Apple’s privacy-first philosophy.

5. Content Delivery and Infrastructure

Video Content Platform: The app will include educational videos from doctors, recorded in a new facility near Oakland, California. These are likely hosted on Apple’s Content Delivery Network (CDN), built on CloudKit, to ensure fast, secure streaming within the Health app.
Backend Services: For any cloud-based processing or data syncing, Apple may use iCloud infrastructure with end-to-end encryption. The AI model updates or external expert content could be managed via Apple’s proprietary servers, potentially leveraging Apple Silicon for high-performance computing.

6. Potential Third-Party Integrations

External Health APIs: The Health app may integrate with third-party health platforms (e.g., MyFitnessPal for food tracking) via HealthKit or custom APIs to enhance data inputs for the AI coach.
Fitness+ Synergy: The workout analysis feature could tie into Apple Fitness+, suggesting integration with Fitness+’s backend for workout content and coaching, possibly using AVFoundation for video playback and analysis.

7. Development Tools and Collaboration

Xcode and TestFlight: Apple’s development team likely uses Xcode for app development and TestFlight for internal testing of the Health app’s new features during the Project Mulberry development phase.
Medical Expert Collaboration: The AI is trained with input from in-house physicians and external specialists, indicating a robust data pipeline for integrating clinical knowledge, possibly managed through custom machine learning pipelines or tools like Create ML.

Challenges and Considerations

Regulatory Compliance: Health advice requires adherence to regulations like HIPAA (U.S.) or GDPR (EU), which may influence the tech stack to include compliance-focused frameworks for data handling and AI explainability.
AI Accuracy: The reliability of the AI coach depends on high-quality training data and robust validation, likely requiring Apple to use advanced model evaluation tools and continuous learning systems.
Scalability: Supporting millions of users globally demands a scalable backend, likely built on Apple’s private cloud infrastructure with Kubernetes or similar orchestration for AI model deployment.

While exact details remain undisclosed, Project Mulberry’s technology stack likely centres on Apple Intelligence for AI, Core ML and HealthKit for on-device processing, Swift/SwiftUI for the Health app interface, and Vision/ARKit for camera-based features. Hardware like the Apple Watch and iPhone, powered by A-series chips and Sensor Fusion, provides critical data inputs. Privacy is ensured through Secure Enclave and on-device processing, with CloudKit and Apple’s CDN handling content delivery. This stack aligns with Apple’s ecosystem, emphasising privacy, performance, and seamless integration.

Nelson Advisors > HealthTech M&A

We work with our clients to assess whether they should 'Build, Buy, Partner or Sell' in order to maximise shareholder value and investment returns. Email lloyd@nelsonadvisors.co.uk

Nelson Advisors regularly publish Healthcare Technology thought leadership articles covering market insights, trends, analysis & predictions @ https://www.healthcare.digital