Your fitness app knows more about you than you think. It knows when you train, how much you lift, which muscles are lagging, how your strength has trended over months, and — if it integrates with Apple Health or Google Fit — it may also know your heart rate, sleep patterns, body weight, and menstrual cycle data.
Now ask yourself: where does all of that go?
The Cloud Problem
The vast majority of fitness applications process user data on remote servers. When you log a set in most workout apps, that data travels from your phone to a server farm, gets processed by algorithms running on someone else's hardware, and the result is sent back to you. Your entire training history lives in a database you don't control.
This architecture exists for understandable reasons. Cloud computing is cheap, scalable, and lets developers run expensive machine learning models without worrying about device capabilities. But it comes with significant trade-offs that most users never consider.
Your data becomes a product. Companies that store user data have an economic incentive to monetize it — through targeted advertising, partnerships with insurance companies, or selling anonymized datasets to researchers and marketers. "Anonymized" is doing heavy lifting in that sentence. Research has repeatedly demonstrated that anonymized health data can be re-identified with surprisingly high accuracy when combined with other publicly available information.
Breaches are inevitable. Every database is a target. Health and fitness data is particularly valuable because it reveals patterns about physical capability, injury history, and health status that traditional data brokers can't easily access. If your workout app gets breached, the attacker doesn't just get your email — they get a detailed profile of your physical self.
You lose control. Once data leaves your device, you rely entirely on the company's policies, security practices, and continued existence to protect it. Companies get acquired. Privacy policies change. Servers get decommissioned. Your training history from three years ago might be sitting on a backup drive in a data center you've never heard of.
What On-Device Intelligence Actually Means
When we say Strenua uses "on-device intelligence," we mean something specific: the machine learning models that generate your workout programs run entirely on your iPhone's hardware. No data is sent to any server. No API calls are made. No internet connection is required.
Here is what happens when Strenua generates a workout:
- Your training history is read from local storage (SwiftData) on your device
- The CoreML model processes that data using your iPhone's Neural Engine
- A workout program is generated based on periodization principles, recovery status, and progressive overload calculations
- The result is displayed in the app and saved locally
At no point does any data leave your phone. The intelligence model itself is bundled with the app at download time — it doesn't need to phone home for updates or instructions.
The CoreML Advantage
Apple's CoreML framework makes this architecture possible. CoreML is designed specifically for running machine learning models on Apple hardware — iPhone, iPad, Mac, and Apple Watch. It leverages the Neural Engine, a dedicated hardware chip in modern Apple Silicon that's optimized for ML inference.
The Neural Engine in current iPhones can perform trillions of operations per second. That's more than enough computational power to run the algorithms needed for workout programming: 1RM estimation, volume landmark tracking, recovery modeling, exercise selection, and load prescription.
A few years ago, running these calculations on-device would have been impractical. The hardware wasn't powerful enough, and the frameworks weren't mature enough. But Apple's investment in Neural Engine hardware and CoreML software has made on-device intelligence not just feasible but practical for real-time applications.
Why This Matters for Health Data
Health and fitness data occupies a unique category in personal information. Unlike your email address or purchase history, your training data reveals deeply personal information about your body:
- Physical capabilities and limitations — what you can and can't do
- Injury history — what has gone wrong in the past
- Health trends — whether your fitness is improving or declining
- Body composition — weight, measurements, and physical changes over time
- Behavioral patterns — when you train, how consistently, what motivates you
This information has value far beyond marketing. Insurance companies, employers, and other institutions have legitimate financial interests in knowing your health status. The more health data that exists in cloud databases, the more pressure there will be to access and use it.
On-device processing eliminates this risk entirely. If the data never leaves your device, it can't be breached, sold, subpoenaed, or monetized. The only person who can access your training data is you.
The Trade-Offs (And Why We Accept Them)
On-device processing isn't free. There are genuine trade-offs:
No cross-device sync. Your training data lives on your iPhone. If you lose your phone, you lose your history (unless you have iCloud backups enabled). We accept this trade-off because the alternative — storing your data on our servers — contradicts our core principle.
Model size constraints. The intelligence models need to fit on your device. We can't use massive models that require GPU clusters. This forces us to be efficient and focused — which, arguably, produces better results than throwing compute at the problem.
No social features. We can't build leaderboards, sharing features, or community challenges without a server. This is fine. Strenua is about your training, not competing with strangers for badges.
These are conscious architectural decisions, not limitations. Every fitness app that offers cloud sync, social features, and cross-platform access does so by storing your data on their servers. We chose a different path.
What to Look For in Any Fitness App
Whether you use Strenua or not, here are questions worth asking about any fitness app:
- Where is your data processed? On your device or on remote servers?
- What is the privacy policy? Does it mention selling data, sharing with "partners," or using data for "product improvement"?
- Can you export your data? If the company shuts down, can you take your training history with you?
- Does it require an account? If you need to create an account with an email and password, there is a server storing your information.
- Does it work offline? If the app requires an internet connection, your data is being transmitted somewhere.
Your training data is your data. It should stay that way.
Strenua processes all workout programming on your iPhone using Apple's CoreML framework and on-device intelligence. No accounts, no servers, no data collection. Learn more about how it works.