Privacy-First by design: what "Data Not Collected" actually means
There’s a quiet trade happening inside most modern apps.
You open an app to get a service — directions, recommendations, predictions — and in return the app quietly collects something from you. Your location. Your behavior. Your routines.
Location-based apps, in particular, sit at the center of what could reasonably be called a data harvesting industry.
That’s why one small line on the App Store page for CommuteTimely matters so much. Under the privacy label, it says: Data Not Collected.
But how can an app that relies on location manage to collect none of it? The answer lies in a simple design decision. Instead of sending your data to the cloud to be processed, we moved the intelligence directly onto your phone.
The Default Model of Location Apps
Most location-based applications follow a similar architecture. Your phone gathers GPS coordinates, which are transmitted to a remote server. The server analyzes the data and sends a result back.
If you send raw GPS coordinates to a server every few seconds, that server quickly learns a remarkable amount about your life. It can infer where you live, work, and sleep.
Rethinking the Architecture
What if the server never needed your location at all?
CommuteTimely is designed so that the intelligence runs directly on your phone. Your device becomes the brain of the system, rather than a thin terminal for a distant server.
Local Storage Only
Your commute contains two extremely sensitive pieces of information: Your home location and your workplace location. In CommuteTimely, these never leave your device.
They are stored locally using Apple’s Secure Enclave—a specialized hardware component designed to isolate and protect sensitive information. Even our company cannot access it. Your commute belongs to you.
Anonymous Traffic Queries
Predicting commute timing still requires external data: traffic conditions and transit delays. The challenge is obtaining that information without revealing your personal route.
We use a technique called disjointed querying. The system breaks your route into small anonymous segments. The server receives fragments like "Traffic on Highway 101 Northbound," not a coherent story about your journey. The system learns about roads, but it never learns about you.
Federated Learning
To improve our models, we use Federated Learning. Instead of sending raw data to the server, the server sends a machine learning model to your device. Your phone trains it locally.
Once finished, the phone sends back only the updated mathematical parameters. These numbers contain no readable information about your commute. We learn collectively without ever seeing anyone’s personal data.
Privacy by Design
Real privacy doesn’t come from promises in a Terms of Service agreement. It comes from architecture. If a system is designed so that it never receives sensitive information, it cannot misuse it.
The Economics of Privacy
Many apps appear free because their true business model lies in advertising networks and data marketplaces. CommuteTimely takes a different path: users pay a transparent subscription for CommuteTimely Pro.
Our incentives align with yours. We succeed when the app is valuable to you, not when we collect more data about you.
A Different Kind of Smart App
Technology should make life easier without demanding unnecessary access to personal information. By moving intelligence closer to the user, we deliver sophisticated predictions without quietly building a database about your life.
That is what “Data Not Collected” actually means. Not just a label, but a system designed from the ground up to respect the boundary between useful technology and personal privacy.