Flipora, a web service that is attempting to rewrite the rules of engagement on the way we discover content on the web — moving away from the now standard “social graph” to an “interest graph” — has gained some critical momentum in recent months.

The web discovery engine constantly analyses its users’ preferences through algorithms that study signals people send out while skimming through content. Using those, the engine “pushes” suggested reading and thus differs from the “pull” paradigm of other search engines where the users key in their queries to pull up information. The premise is that the web is just too big to just browse based on keywords or what one’s friends recommend.

Indian connection

What makes the story even more exciting is that it is founded by two Indian graduates from Stanford University.

These are exciting times for Jonathan Siddharth and Vijay Krishnan, students from the class of 2007-08, who are seeing their Silicon Valley start-up gather steam. Their service reached its 10-millionth user mark recently, acquiring users from over 200 countries. The speed at which the company is growing is also cause for celebration — it took Flipora 45 days to go from the 8-million to the 9-million mark but just 25 days to go from the 9-million to the 10-million mark.

The start-up has so far raised funds close to $3.9 million through two rounds of funding from prominent Silicon Valley firms such as DFJ (backers of Skype, Paypal, Hotmail, Baidu, SpaceX, etc.), Labrador Ventures, Band of Angels, Amidzad Partners and Steve Oskoui (Smiley Media). It now has 20 employees and is headquartered at Sunnyvale, California, in Silicon Valley.

At their current pace, and based on their revenues from Google AdWords, Mr. Siddharth and Mr. Krishnan hope to break even and turn in the profits the first quarter of next year. “If things go as good as they are right now, we might never need any further funding,” says Mr. Siddharth hopefully.

Discovery engine

Flipora started out as Infoaxe in late 2008, a service that in its initial stages came across as a “complete registry” of a user’s online activities. Using a web browser plugin, the entire history of one’s browsing would be ranked and stored, just to answer the question — “What was that site I fleetingly looked at when I was browsing two months back?”

Re-finding queries

A 2007 study by a group of students from the Massachusetts Institute of Technology found that nearly 40 per cent of all search queries were re-finding queries. By banking on a user’s entire web history to help search for particular data, relevance of results improves dramatically.

But in July this year, Mr. Siddharth and Mr. Krishnan re-branded their service as ‘Flipora’ and used the same underlying logic to help users discover new websites based on the interests that their browsing habits showed.

Today, users can register with Flipora with details about their areas of interests and start browsing recommended content from like-minded users the world over. The service has been adding several categories and has also incorporated certain algorithms that make the process both fun and smart. In some ways, this is the kind of discovery of content that has faded away because of the impact of Facebook and Twitter.

Refining categories

In a few ways, Flipora comes across as being similar to Pinterest and Quora. But they are some key differentiators too. Pinterest is more picture-oriented and Quora can come across as a tad geeky. Flipora is more about articles and content recommendations. Mr. Siddharth says they have been refining the categories.

Flipora also has an option for its users to follow certain ‘recommended curators’ helping users understand what sort of interests they might share with them. Popular singer Seal has expressed interest in being featured as a recommended user with the service.

As the Flipora discovery engine gets closer to its tipping point, Mr. Siddharth and Mr. Krishnan are planning the big push forward that should come next year. Their idea is to make Flipora a third-party discovery plugin that would be incorporated into sites in a way similar to how “social graphs” of networks such as Facebook and Twitter are plugged into sites today. They call their project the “interest graph.”

“Won’t it make more sense if the e-commerce website you are browsing throws up recommendations based on your interests rather than your friend’s interests,” asks Mr. Krishnan.

More In: Technology | Sci-Tech