Our Friday CEO Spotlight series continues with today’s guest Venky Harinarayan, Co-Founder of Kosmix.

Venky has a PhD from Stanford University in Computer Science and has been been featured in the Wall Street Journal, Salon.com, India Today, CNET TV and on the cover of Business Week(Intl. Ed.). Together with his Kosmix co-founder Anand Rajaraman he created and co-founded Junglee, the first shopping search engine which was acquired by Amazon.com in 1998 for $250 million. At Amazon.com, they created the search and marketplace business, a key part of Amazon today, accounting for 30% of gross merchandise sales.

Venky is also a founding partner of Cambrian Ventures. The firm has invested in such Internet companies as Efficient Frontier, Kaltix (acquired by Google), Neoteris (acquired by Juniper), Transformic (acquired by Google), TheFind, and Aster Data. He serves as investor, advisor, or board member to several technology start-ups.

Venky and Anand founded Kosmix in 2004 with a mission to “Build the unofficial homepage for every topic on the web.”

ASE: Hi Venky! Welcome to Alt Search Engines.

Venky: We’re excited about this opportunity.

ASE: I’ve spent a few enjoyable days playing with Kosmix. Good stuff!

Venky: Thanks. We’re doing something hard and are still in our early days.

ASE: So, what inspired you to found Kosmix? Was there one specific search that you conducted that was just very unsatisfying?

Venky: Our inspiration really was the Web. We started with this belief that the Web is like the Library of Alexandria, a repository of all known human knowledge. And it felt like search was but one window into this repository.

ASE: Where was all this knowledge before the web? I’ve been wondering lately…

Venky: That’s a fascinating question. Our vision was there had to be other ways to access this knowledge. Think about medical information — let’s say someone you knew was diagnosed with high cholesterol. What did you do pre-Web?

ASE: It was so labor intensive!

Venky: I think you asked someone. Maybe went over to your library?

ASE: You asked someone, went to the library, harassed your doctors…

Venky: Remember those?

ASE: I know. It is so empowering to know you have all of this access at your fingertips anywhere in the world. All you need is a question…

Venky: I think the cost of accessing information was so high, that it was practically unavailable for most things. Consider stock quotes — we had to call brokers.

ASE: Where are those brokers now?

Venky: I think they’re out of jobs… So while I think the Web and search have changed the landscape, there is so much more that is possible

ASE: Do you feel that up until now web search has been in its, so to speak, beta phase?

Venky: If you look out 10 years, it’s hard to see information still being accessed the same way. I think there will be two big trends: The first we’re starting to see is people no longer see the Web as a collection of Web Pages. Search still does. Consider planning a trip to New Zealand. There are sites with factual data, there are review pages, there are blogs, videos, slideshows, social network groups… Each one means something different to a user. There are widgets like a weather widget, webcams etc… All of which are valuable to me. So how do you rank a blog article against a review? You can’t.

ASE: Well… there’s the issue of trusted info from publications that have standards and some dude named Joey’s blog.

Venky: Yes, so would you show 10 results of really trusted information, no opinions? Search today is one dimensional - everything is a web page, we order results 1 to 10. What we’re saying is the world will move to dimensions. Everything has a place and you can’t compare an opinion to factual data and try to rank them. What we’re saying is if you asked an editor who could look at the entire Web and tell you what you should know about New Zealand, what would they do? How do we do that?

ASE: How does Komix decide which sources are “Trusted Sources?”

Venky: Trusted Sources — information that comes from Medical Organizations and other non-profits.

ASE: Ok… Sorry for interrupting you midstream - back to New Zealand.

Venky: I didn’t take a vacation this holiday season, guess I need to go somewhere! I was in New Zealand a couple years ago, and it was really quite spectacular.

ASE:I have family on the South Island- Otago.  I’m waiting for an invitation.

Venky: You do? Fancy that — that’s some coincidence, considering NZ is smaller than the bay area. The South Island is amazing, especially Milford Sound.

ASE: It must have been amazing.

Venky: Yes.  Ok, back to our regular programming! So I think info finding moves to dimensions. This is driven largely by the explosion of user-generated content. The second trend, I see is a little more futuristic.

ASE: Web 2.0

Venky: Yes… emergence of the Subjective Web if you will. The Omgili guys talk to this well… So the second trend I see is a little further out, though we are scratching the surface at Kosmix. Right now search is about Web intelligence –> People. Big change will be Web Intelligence –> Programs.

ASE: Please elaborate…

Venky: So think of the PC — today you hardly use the Operating System (creating files, etc… You operate on programs. Application.

ASE: OK

Venky: I think the Web will evolve the same way. For example, if you think about the Web it has so much medical information out there. Why don’ t we have a diagnostic utility that can have an interactive dialog with us?

ASE: A smart web that doesn’t just contain info but shares it in context, so you’re not searching for the info, the info is being fed to you-all you need is the question.

Venky: Sort of. I am saying we need to channel the Web’s intelligence to programs. Programs can do much more processing than humans can, and you can get some unbelievably valuable apps. Think about trying to do fancy spreadsheets without Excel. At Kosmix for example, given a query like diabetes, we automatically infer the symptoms, drugs, treatments. We mine the web for this info and feed it to programs

ASE: And Kosmix does this via a crawler- it is not human search, correct?

Venky: http://righthealth.com/Health/diabetes/-od-connections-s. These are the connections we mined for diabetes from the Web. Yes, we’re crawler based.

ASE: That’s amazing… Very helpful. I also looked up canine renal disease since my dog has this problem and found more/better information than on Google.

Venky: That’s not good… I mean for your dog (or Google) - I have a dog too. 

ASE: He is a 13 year old poodle.

Venky: I have a 10 year old cocker spaniel. Can’t live without him!

ASE: What’s his name?

Venky: Junglee. The name of my first company. We bought him the week Junglee was acquired by Amazon. We were very scared when the Chinese food scare happened.

ASE: Were you born in India?

Venky: Yes. Bombay, now called Mumbai, lived in Madras, now called Chennai. I left before they changed names, so I keep getting my dad upset by calling Chennai Madras…

ASE: I was born in Odessa, Ukraine but when we emigrated it was still called Russia, so there is the name confusion too- am I Russian or Ukrainian?

Venky: That’s former USSR right?

ASE: Yes. I have never been to Chennai, but I would love to go to India. I have a good friend from Bengal who always invites me.

Venky: Yes, go in December.  BTW, growing up we used mostly Russian books.

ASE: Interesting- and I never read Russian books, but many Indian ones. Ok, back to business. When did you come to the U.S.? And was it alone or with family?

Venky: I came alone – 1988. I went to UCLA.

ASE: And then Stanford.  Is that is where you met Anand?

Venky: Yes, I met Anand at Stanford, we were both graduate students there, and we had the same advisor. And collaborated on a couple of papers pre-Junglee.

ASE: I have read about you guys and Sergey Brinn, etc. Must have been an interesting time to be there!

Venky: Yes, it was a lot of fun.

ASE: And then to be at Amazon in 1998, did you always know Amazon would be as big as it is today?

Venky:  It was quite interesting.  Amazon, when we went there, they used to think of themselves as this little start-up. Coming from the outside we had this perception they were this large company. Of course now, they’re truly a large company.

ASE: Wow, that must have been something. So both you and Anand went to Amazon together?

Venky: Yes we did. In 1998

ASE: You’ve been a team for a decade.

Venky: Yes, scary, huh?

ASE: So what do you think will happen to all of these alternative search engines in the next ten years? Do you think they will consolidate, be bought out?

Venky: Here’s the big problem for all these search engines — if you build a great search engine for health, that’s awesome, but health is like 5% of Web Searches online. Plus it is episodic, you don’t search for health everyday. So how do I get my users to keep coming back? The way we’ve addressed that is by working in the context of the existing ecosystem. We’re happy if our consumers find us starting at Google, Yahoo, MSN. Our vision is to build the best “home page” for a topic. So we see our product more like Wikipedia than Google. The results are really starting to show. In December we were in the top 5 health sites online and in the top 15 auto sites online. Righthealth and Rightautos are our health and autos properties.

ASE: What about the other guys? It is such a crowded space.

Venky: I think it’s hard if you want to displace the current search engines, though I’d say the rewards if you do are enormous. The question to ask is how do I make a difference to a consumer who’s looking in health, autos, travel, whatever, and you quickly realize you need to go where they are.

ASE: Do you pay for any content?

Venky: We’re fully focused on delivering value to our consumers, and have found in a few cases that they’d like to get baseline content from us. So we do pay for content in few cases. It’s a very small number.

ASE: How big is your team, and, do you outsource to India?

Venky: We’re around 50. We keep options open around outsourcing, so we do have a small footprint in India, but it’s something that’s pretty small at this point. The issue with outsourcing and a startup– outsourcing is usually about cost efficiencies, not time to market. So you need to be careful because in a start-up it’s mostly time-to-market.

ASE: Right. On the topic of India.  Are there any interesting Alts coming out of India?

Venky: I have seen a couple — guruji.com and some activity in local. India still feels like the battle is around the e-commerce plays, like travel, jobs, etc…

ASE: Well, Venky, thank you so much for your time, I’ve enjoyed chatting with you!

Venky: Good talking with you too.

Natalya Murakhver is a freelance writer/PR consultant based in New York City.

AltSearchEngines is looking for a sponsor for this feature. Your company’s logo, tag line, and link would appear here after evey Friday profile. Please email Charles@ReadWriteWeb.com for more information.

This entry was posted on Saturday, January 12th, 2008 at 7:35 pm and is filed under CEO Views. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Sphere: Related Content