Zero-Egress In-Browser Document Processing Systems Using Web Workers
The paper introduces HandleMyFile, an edge-native, zero-egress document processing system that executes entirely within client browsers via Web Workers to eliminate network latency, reduce costs, and ensure data privacy, demonstrating performance that significantly outperforms remote cloud services while guaranteeing zero outbound data transmission.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the modern digital world, we often treat our computers as remote control panels for vast, invisible data centers. When we need to edit a document, merge files, or remove sensitive information, the standard practice is to upload the file to a cloud server, let a powerful computer there do the work, and then download the result. This convenience comes with a hidden cost: every time a file travels to a server and back, it crosses a network where it can be intercepted, logged, or accidentally left behind. Furthermore, the time it takes for the file to make that round trip depends entirely on the speed of the internet connection, creating delays that can frustrate users and slow down business.
A different approach has emerged, one that shifts the work from the distant server back to the device in the user's hand. This method relies on a technology called a "web worker," which acts like a dedicated, silent assistant running inside the web browser. Unlike the main browser window, which handles the visual interface and can freeze if asked to do too much at once, a web worker operates in a separate, isolated thread. It can perform heavy calculations without interrupting the user's view of the page. By combining this isolation with strict rules that prevent the browser from sending any data out to the internet, researchers have created a system where documents are processed entirely within the user's own computer memory, never leaving the device.
Khairan Noor Fadhlillah, a researcher at Universitas Ary Ginanjar, has developed and tested a system called HandleMyFile to prove that this local approach is not only possible but superior to the traditional cloud method. The system is designed to handle complex tasks like merging, splitting, compressing, and sanitizing PDF files. The core idea is simple yet radical: keep the data local. By enforcing a digital lockdown that blocks any attempt by the software to send data out over the network, the system ensures that a user's private documents never travel across the internet. This eliminates the risk of data breaches during transmission and removes the waiting time associated with uploading and downloading large files.
To verify that this local processing actually works as well as the powerful servers used by major cloud companies, the researcher conducted a rigorous test involving 490 real-world documents. These files ranged from tiny, single-page notes to massive, multi-hundred-page reports, all drawn from a standard collection used to test PDF software. The study compared how long it took to process these files on a local computer versus how long it would take if they were sent to a server. The results showed that the local browser-based system was not only faster but also mathematically identical in its processing power to the server-side engines used by professionals. In fact, on a local network where the internet speed is effectively infinite, the browser system was nearly twice as fast as the server version, simply because it avoided the overhead of moving data back and forth.
The advantage becomes even more dramatic when the internet connection is slower. The study modeled real-world conditions, such as a typical home Wi-Fi connection or a mobile 4G network. In these scenarios, the local system outperformed the cloud server by a factor of more than ten on Wi-Fi and by nearly thirty-eight times on mobile networks. This massive speedup occurs because the local system does not have to wait for the file to travel to a distant server and back; it does the work instantly in the computer's memory. The research confirmed that the browser's ability to handle these tasks is not a compromise; it is a genuine alternative that matches the performance of high-end servers while offering a significant speed advantage in everyday network conditions.
Beyond speed, the system addresses a critical security concern: the leakage of hidden information. PDF files often contain invisible layers of data, such as the name of the author, the software used to create the file, or timestamps of when edits were made. Even if a user deletes the visible text, this hidden metadata can remain, potentially exposing confidential details. HandleMyFile includes a specialized engine that strips away these hidden descriptors with extreme precision. In the tests, the system successfully removed 99.71% of all metadata fields from the test documents, leaving behind a clean file with no trace of its history or origin. This level of sanitization happens entirely within the browser, ensuring that no sensitive information is ever sent to a third party for processing.
The security of this approach was further tested by subjecting the system to 100 different types of malicious or broken files designed to crash software. These included files with corrupted structures, infinite loops, and massive amounts of hidden data intended to overwhelm the computer's memory. The system handled every single attack without failing. When a corrupted file was encountered, the isolated worker thread simply stopped and restarted, preventing the crash from affecting the main browser window or the user's computer. Throughout all these tests, network monitors confirmed that zero bytes of data were ever sent out from the user's device. The system remained completely silent on the network, proving that the documents never left the local environment.
This work demonstrates that the future of document processing does not necessarily require sending files to the cloud. By leveraging the power already present in modern web browsers and isolating that power within secure, local threads, it is possible to perform complex, enterprise-grade tasks with greater speed and absolute privacy. The study shows that for the vast majority of users, the most secure and efficient way to handle sensitive documents is to keep them right where they are, processing them locally without ever crossing the network boundary. This shift from a reliance on remote servers to a model of local execution offers a new standard for privacy and performance in the digital age.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.