When you start working with Rails, you know it's powerful - it does a lot of heavy lifting, and where it leaves off, RubyGems usually pick up. But have you ever wondered if you were missing something?
We did. Sure, you can browse RubyGems.org. You can manually inspect Gemfiles on GitHub. But now there's a way you can automatically analyze the first ever "gem network" using heuristics for related, similar, and complementary Gems.
Simply upload your Gemfile and we'll do all sorts of crazy math to predict a few Gems in which you may be interested.
When you upload your Gemfile, we parse it to find all
For each Gem listed, we calculate a list of recommendations based on other
Gemfiles we've seen and then we combine those lists to produce the final
set of recommendations.
In the background, we process these uploaded Gemfiles* to make our algorithm even smarter.
*Paranoid? We understand. But don't worry - when you upload your Gemfile we discard the actual text file immediately upon extracting the listed Gems. We then use only a fingerprint of that derived data to identify it in our systems. The actual text of the Gemfile - including all the sources, groups, comments, etc. - stays in RAM for but a fleeting moment. Not only is there no analytical reason for us to keep it, we simply don't want to be responsible for anyone's data which could be sensitive.