NordPass password manager Machine Learning-Enhanced Autofill
NordPass
Category
AI/Machine Learning Innovation of the Year
Entry Description
This year NordPass has introduced a major improvement to one of its key features — autofill. Its form-detecting algorithm became more efficient by introducing machine learning. Before the update, developers had to manually improve the form-detecting algorithm.

The part of the autofill algorithm responsible for detecting the type of form now operates with vastly improved performance compared to its previous static iteration. The algorithm is now trained using artificial neural networks to be as accurate as possible. According to NordPass developers, the current accuracy of the algorithm is 99.6%.

The new update has two main benefits. The first, more accurate form detection, leads to more accurate autofilling, which results in an enhanced user experience. The second benefit is that machine learning will give developers more time to further improve other aspects of the product.

It is important to note that the learning process of the machine learning algorithm is not intrusive. It doesn't take place on users’ devices. It is done by developers and is released with each NordPass update. Users don’t need extra resources on their devices. The learning process takes place in a controlled environment until it is ready to be shipped out. The machine learning process is not entirely autonomous. Before an update, developers collect lots of data, train the algorithm, and get a small file, which then detects forms and therefore improves the entire autofill process.