The easiest way to add search to your documentation

get it
#2 Product of the DayDecember 24, 2015



You need to become a Contributor to join the discussion - Find out how.
Nicolas DessaigneMakerHiring@dessaigne · Co-founder & CEO, Algolia
Hey hunters, As the year winds down, the team at Algolia would like to help you celebrate with our 2015 holiday gift: DocSearch. With DocSearch, we’re making it easier to search within your documentation, and we’re doing it for free. As in, you just fill out the form, and we send you the code to implement DocSearch within your documentation. You don’t need to create an account or send us any money! Check out some existing implementations like and and let us know if you’d like DocSearch for your own documentation :) Happy Holidays!
Tom Benattar@tombenattar · Co-founder @PixelMe
Really nice @dessaigne! Congrats 😄
Ingo Radatz@llabball · CTO, R&D at HSH / Berlin, Germany
Thx for the gift 👋. Could you please explain what technically enables the search? Is it client-side/server-side, do you use a tokenizer ....
Sylvain UtardMakerHiring@sylvainutard · VP of Engineering @ Algolia
@llabball The search is powered by the Algolia search engine, a hosted search engine. Under the hood, it's C++ code hosted in an NGinx module. The small JS code snippet that we provide you calls the Algolia API at each keystroke, to retrieve the most relevant results. You can read about it here: or
Ingo Radatz@llabball · CTO, R&D at HSH / Berlin, Germany
@sylvainutard Ah ok, a server-side search. Every keystroke means you expecting to being able to handle a lot of requests in-time. Can you crunch some numbers with us? What's your capacity in terms of concurrency, avg response time ...
Sylvain UtardMakerHiring@sylvainutard · VP of Engineering @ Algolia
@llabball Oh yes, that's a bunch of API calls a month :) Last month we processed ~8B queries out of a total of ~30B operations done on our API. On average, search queries are processed in <15ms and with our Distributed Search Network ( we're confident about spending <60ms on the network wherever you're on the planet: realtime search \o/ Some of our customers have 1 big index with up to hundreds of millions of objects; other have up to thousands of small indices, it scales pretty well :) For DocSearch, we've set up an Algolia account replicated in 10 regions around the world to provide the best search experience to all your end-users.
svetlyak40wt@svetlyak40wt · Founder of
@sylvainutard @dessaigne is it possible to use Doc Search in a project like AllMyChanges or there are some license limitations?
Ingo Radatz@llabball · CTO, R&D at HSH / Berlin, Germany
@sylvainutard wow. Thx for the insights.
Jules@scomopolitan · AFILIZA
Inspired by Uber for the design? Looks like a powerful product
Kevin GrangerMakerHiring@kevin_granger · UI-UX Devsigner, Algolia
@scomopolitan hehe uber for doc, ok :)
Nasser@coosmicboy · Front-end Developer - React/Redux
This seems promising!