Even a brand new computer can be slowed down by a particular combination of hardware or software, and it’s difficult to figure out what’s wrong. Soluto, a startup based in Tel Aviv, Israel, that launched this week, hopes to offer users advice on how to avoid this kind of slowdown.

Soluto’s software runs in the background on a PC and is designed to detect problems that slow a machine down as well as solutions that speed it up again. The idea is to collect this information and use it to recommend software and hardware fixes to other users. “We know when you’re frustrated, we know what causes it, and if you do something smart to your PC, we learn from it,” says Roee Adler, chief product officer for the company.

Soluto’s first product, launched this week at the TechCrunch Disrupt conference in New York City, is designed to address problems that slow down a computer as it starts. The free version of the software simply observes what’s happening on the machine and recommends fixes to speed up the process. The company also sells a premium version that performs fixes automatically. Soluto hopes that this product will appeal to less tech-savvy consumers and to small businesses and enterprises. This product collects a lot more information about what’s happening on a user’s PC, which it stores in Soluto’s “PC Genome” database.

Adler explains that the company has spent two and a half years working on algorithms that can recognize when a user is frustrated with his or her PC. The software learns from how people deal with these problems in order to recommend changes that can help other users. Soluto’s software communicates with a Web server that stores statistical data that’s used to interpret what’s happening on the machine.

The software observes events such as repeated mouse clicks–which may indicate that the user is not getting a response–and sudden shifts in how the PC is using resources, which may mean that programs are fighting over resources. Users can also manually inform Soluto that the computer is experiencing difficulties, and this causes the system to pay particular attention to recent events on the machine.