Software is transforming the world, but we can’t write it fast enough. There aren't enough software engineers in the world, and this is an important limit on growth worldwide. So, how can we solve this problem?
Increasing developer productivity is the key. We're using machine learning to eliminate the repetitive parts of programming and make developers much more productive.
Kite released Line-of-Code Completions for Python in January 2019 with advanced statistical modeling to incorporate context and complete full lines of code in real-time.
Our stack
Kite is cross-platform desktop software that supports macOS, Linux, and Windows. Our desktop client is mostly written in Go, with a bit of C for a few cross-platform components. We integrate with a variety of editors and IDEs and build those integrations with languages including Javascript, Python, Java, and Vimscript. Each editor plugin “talks to” the Kite desktop app, which is where the intelligence lives. The desktop app also bundles Tensorflow, which runs our machine learning models.
