The open source Elasticsearch is being brought into many areas of HSLynk, to speed it up under load. First, we have added Elasticsearch to our web client search APIs.

We want your human services apps to be very responsive, so we've been speeding up different parts of HSLynk. Elasticsearch is a leading enterprise search system, and its incorporation into HSLynk represents our commitment to making your human services apps run faster on our platform.

The initial search for a particular client is a central activity to community information exchange, and we've optimized it for you.  It wasn't easy, because the list of client names in HSLynk is always being updated from many parts of the system, whether it's HMIS or resource matching, so it always has to be listening for published changes from our "pub/sub" engine.


To try out our implementation of Elasticsearch, sign up for a developer account at, and then use your app to call our client search API.  The video below shows a small test of our old search running side-by-side with our new Elasticsearch.

The tests in this video were conducted by a client external to HSLynk, which represents the full round-trip web request time. 

Next, we plan to add Elasticsearch to our client dedupe algorithm.  Stay tuned for that in the coming weeks...