Kite - AI Powered Auto Completion for Python
Written by Nikos Vaggalis   
Monday, 09 September 2019

Productivity is not just associated with saving keystrokes but it comes from making smart suggestions too. This is something that Kite does with its new AI-powered Intelligent Snippets.

I discovered Kite back in 2017 and presented its capabilities in Kite - Smart Copilot For Programmers. Today's news is that it's been updated with AI capabilities, which I'm going to go through after a quick refresher about what this tool is and what is can do for you.

Kite is a plugin for your IDE that acts as a coding "copilot", likened to the buddy with whom you do peer coding. Kite integrates with your favorite editor and monitors your typing in order to display highly relevant information. For example, Kite unveils Python packages or function signatures as you type:

together with one-click relevant documentation:

and popular code patterns based on how other people used them:

kite-popularpatterns

Kite also spots spelling mistakes and syntax errors and offers to fix them for you. What's even more interesting is its context sensitive awareness as it can predict what you are planning to do and provide hints.For example, let's say that we make a request to a URL:

response=requests.get('http://....');

Kite anticipates that what you typically do after retrieving such an object is to parse it and therefore suggests using a "response.json" call.

 

Since 2017 a lot more useful features have been added, some being milestones in the tool's history. One such milestone is this September's addition of AI powered Intelligent Snippets, that are generated in real-time based on the code patterns Kite finds in your codebase. This is where Machine Learning comes into play:

Kite first indexes your codebase and learns how functions are commonly used. Then when you call a function, Kite suggests snippets for that function to easily complete it. Kite’s autocomplete still suggests completions for each argument, too. 

 

Adam Smith, CEO of Kite notes that :

Intelligent Snippets work like a “fill-in-the-blank” experience for developers so they don’t have to remember how to complete every function call they need. Whereas devs usually copy-paste snippets manually from the internet, Kite’s machine learning engine now automates this workflow

And we all know how much devs love copying and pasting code found online, now even empirically proven as part of the research we recently reported in Can Regular Expressions Be Safely Reused Across Languages?

Intelligent Snippets also work on functions that you have defined yourself, which is why it scans your code in the first place. In short Kite reduces both key strokes and the time taken to lookup documentation. And the ability to copy/paste code snippets certainly makes for a much more productive programming experience.

Pythonistas are the lucky ones, since Kite's core functionality is available only for Python. That is only for the time being, however, since  serious effort is being devoted to ports to other programming languages. As far as IDE availability goes, Kite integrates with Atom, PyCharm, Sublime, Vim and VSCode.

So can you put a price on such coding "friendship"? No actually, since Kite won't cost you a penny. It is completely free with no strings attached. The team just focuses on creating a product that developers love to use. I'm sold!

More Information

Kite announces Intelligent Snippets for Python

Related Articles

Kite - Smart Copilot For Programmers
DeepCode Gets Cash And Opens Free Tier

 

 

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on, Twitter, Facebook or Linkedin.

 

Banner


AI Learns To Solve Rubik's Cube - Fast!
31/08/2019

Yes, this is another "AI masters a game or puzzle" piece of news, but this one is surprising. Having learned to solve a logical problem, AI turns in a performance that typically not only beats a human [ ... ]



ReduKtor - Auto Debug For Kotlin
18/09/2019

No, this is not an AI approach but a good old down to earth engineering approach to isolating bugs. It's a technique that you could well apply manually, but as an automatic tool - even better.


More News

 

graphics

 



 

Comments




or email your comment to: comments@i-programmer.info

Last Updated ( Monday, 09 September 2019 )