Friday, October 1, 2010

LinkedIn Signal

A few days ago, LinkedIn Signal debuted at Techcrunch Disrupt. Within hours, tweets, blogs, articles spread like wildfires across various mediums on the web. It was exciting to be part of the engineering team behind it.





Here is an encouraging article written about Signal:
http://www.thedailybeast.com/blogs-and-stories/2010-09-29/linkedin-signal-combines-twitter-with-the-resume-site/

Signal is a contextual social search application that leverages the realtime LinkedIn Share/Twitter stream and LinkedIn profile information. In a way, we were able to classify the realtime stream based on who the "Sharer" is, and thus tagging it with structured information.

Being a data product, having clean and abundant data is essential. With combination of LinkedIn and Twitter data, we were in heaven in building a revolutionary product: In a few clicks you are able to answer the following questions:
  • What are HP employees saying about Mark Hurd?
  • What is Jerry Brown Campaign saying about Meg Whitman?
  • What do MIT students think about Scala?
  • What are the IT professions living in SF-Bayarea talking about Java?
  • What are people saying about LinkedIn Signal in the last hour?
  • ...
These are only a few examples of the type of insights you can get from the realtime stream that has been enriched with context and precision.

Furthermore, you can also discover trending articles you should read based on your query and selections. We were suggested to read the Techcrunch article about AOL buying Techcrunch hours before any news source picked it up via Signal, that was powerful!

I have written an under-the-hood technical post on the LinkedIn SNA blog, so instead in this post I would like to talk about the development process for which Signal was created.

A few months ago, one of our rock star engineers Nick Dellamaggiore wrote a data stream that merges LinkedIn shares, the Twitter stream for bounded LinkedIn accounts, LinkedIn profile, and derived LinkedIn member information, such that any LinkedIn developer can consume and build interesting applications from.

For a few of us that have been interested in making sense of semi-structured data in realtime, having access to this data stream straight from our development boxes is like mice trapped in a cheese factory, we were excited!

Then the stars aligned some more, a ridiculously awesome application developer Alejandro Crosa took our search library and presented us with a beautiful application, an incarnation of Signal appeared before our eyes. And a team was formed, led by our product counterpart: Esteban Kozak.

For the next couple of weeks, we went nuts with features -> performance -> more features, and worked through weekends and evenings, we were in startup-mode.

Judgement day came, we presented Signal to our CEO, Jeff, and our VP of Products, Deep. Being internet veterans, immediately saw the value, provided feedbacks and demanded execution! Yes, this is how we roll in the Silicon Valley!

We got the entire company excited, people from different groups and organizations pitched in: e.g. Operations, Engineering, Design, Products, Marketing... and a beautiful thing happened: Collaboration!

On September 29th, we showcased Signal, our baby!

Building Signal brought out the essence of the Silicon Valley, the birth places of Google, Yahoo!, Facebook, Twitter etc.

Sunday, January 31, 2010

LinkedIn Search Talk - SDForum

The past Wednesday I had the pleasure of giving a technical talk at SDForum on LinkedIn Search.




This talk came about a month after the 100% rollout of LinkedIn Faceted People Search. See blog by Esteban Kozak.

In this talk, I talked about the LinkedIn search infrastructure that hosts various LinkedIn search properties, e.g. people search, news search, job search etc.

The main features we built are:
  • realtime indexing/search
  • streaming/live update
  • faceted navigation
  • section search
  • distributed index partitioning
The slides provide a glance through how we built these technologies through the following open source projects we built/working on:
  • Zoie - realtime indexing update system
  • Bobo - faceted search engine based on Lucene.
  • Sensei - distributed realtime faceted search system.
Some of the notable attendees are:
along with representations from companies such as Google, Apple and VMWare etc.

I am glad to have learned different uses for search technology and hope the technologies we have built to be helpful in different areas.

For learn about our team at LinkedIn and see other open source projects we are working on, visit http://sna-projects.com/.