company · 2026-11-02 · 2 min read
Introducing GDPad: privacy starts on your device
Perfectory is building GDPad, a desktop security solution that uses local AI and rules to detect and mask sensitive information before requests reach external AI providers.
We’re building GDPad — a desktop data security solution that protects sensitive information before it reaches external AI providers.
AI tools are becoming part of everyday work. People use them to draft documents, review information, write code, and solve problems. In the process, a request can contain more than the task itself: a customer’s contact details, confidential business information, or content from an internal document.
That moment before information leaves the device is where we’re focusing our work.
GDPad operates at the system level, intercepting requests before they reach external AI providers. It uses local AI models and rules to detect and mask sensitive information on the user’s desktop.
The purpose is straightforward: help people use AI while reducing the risk of unintentionally sharing sensitive data.
Why we’re building locally
The architecture behind a security product shapes how it handles the information it is meant to protect. With GDPad, we chose a local-first approach, placing detection and masking on the device where the work happens.
This gives us a clear point at which to apply protection: before an outgoing request reaches an external AI service.
Our goal is to make that protection fit naturally into everyday workflows. AI adoption should be safer without creating unnecessary friction for the people using it. That principle guides how we think about the product, from its architecture to its desktop interface.
We believe privacy and data security should be built in from the start. For GDPad, that means considering how sensitive information is handled as part of the product’s foundation.
An early look at what’s next
GDPad is being developed by the team at Perfectory, and we’re preparing a prerelease version.
We’d like to hear from people and teams working with AI: what information needs protection, where concerns arise, and what would make a desktop security tool useful in their day-to-day work. Those concrete scenarios will help guide the product as it develops.
If you’d like to try an early version, ask a question, or discuss your use case with us, visit gdpad.io. You can request prerelease access or book a call with our team.
Privacy starts on your device. That’s where we’re building GDPad.
Planning an AI automation initiative? Read the AI automation POC guide.