Hacker News new | past | comments | ask | show | jobs | submit login
Ark (YC W12) launches a modern people search engine (techcrunch.com)
132 points by priley on March 25, 2012 | hide | past | favorite | 53 comments



Am I the only one who feels that any kind of search engine that requires creating an account is a non-starter?


Great point. We'd love to completely open up and allow you to search without authenticating, but part of the app allows you to setup a private index of all the private data of your Facebook friends. We need to scale that out a little. But you are right.


Don't underestimate the power of stalking to encourage people though.


another data broker to add to the list on safeShepherd.com


:) Ha.


Hi guys, really nice work.

I am really into the tip of the iceberg idea for the website.

One question to the developers that I am sure it is in many people's minds: 1) I sign up 2) U use my login credentials to access the apis 3) U print the results on my screen 4) U save the "business card" of the person I searched for without his consent. And now the million dollar question. Do you for any reason give FULL access to this business card to: a) a person that has partial access to my contact for example linkein connection but not facebook friend? b) to any other person or company that has no connection whatsoever, now or IN THE FUTURE?

In other words do you provide details that are not publicly available to people that have no connection?

Sorry for the long post but I am a little bit concerned about privacy...


Great question. So, there are two parts of Ark:

1) Massive index that spans across all the world's social networks. This publicly accessible data is normalized, cleansed, etc.

2) You can make a "private" index of just your friends. This never goes outside of your own private index. That data remains there and is not added to the massive index. It is incrementally updated, so if people are no longer your friends (bummer!) they are no longer in your private index.

No one gets full access. Otherwise we lose all our integrity. I hope that helps.. and thanks again for the question.


question to the founders - how did you get this TREMENDOUS domain name?


I'd be fascinated to hear this story too. I'd think that domain would be worth more than most ycombinator startups could afford.


The story is actually a great one, and will be told someday. I'll save it for the book that is turned into a movie about how I made Ark just because I wanted to be as cool as a crew guy.


priley must have amazing negotiation skills.


Or patience. If you linger around domain auction platforms and forums enough and are willing to invest you can get lucky on a relatively consistent basis.


It was a little of both. ;) Hey... everyone... I just wanted to thank you for all your comments about our new search engine! This community has such great insight and I just really appreciate the support with Ark. It has been an exhausting 45 days of coding and designing, but encouragement like this makes it all worth while.


"First name Emily + birthday in March + Studied at NYU = That ridiculous girl from New York you met during her birthday party at the bar last night"

sounds like a great addition to my cyber stalker toolbag. These cross-network profiling is really one of these areas which haven't been covered that much yet, except maybe by intelligence services. Looks like an interesting project to watch.


Google and Facebook can’t help you find which of your friends are single, or live in New York and like Radiohead...

Really? I think that it can.


I think what he means which of your friends are single AND live in New York AND like Radiohead.

Search only gets interesting when you learn how to power search, and Ark is that power people search.



What ever happend to greplin? weren't they also kind of going to solve this? They were in the news a ton until they weren't.

But yeah, we need this. Even Facebook's search has been nerfed compared to where it was years ago. It is a colossal pain, all the data is there but no one wants to organize it


Greplin is a personal search system, isn't it? Its corpuses are private.


Greplin is awesome and Daniel's team is super impressive. But they focus and index on all your private data like messages, emails, etc. We are trying to improve the 30% of internet searches online that are people-related. I love how you say Facebook's search has been nerfed.. I may borrow that for the Demo Day preso. :)


:) please do. It was. It was always a bit limited but there used to be a decent set of controls. Filter by gender, age, likes, institution, even dating status I believe. Now practically nothing. I think name and possibly sub filter by group. That's it. It's pathetic, and pretty much useless for anything other then finding people in your list by name faster than scrolling down the page.


The companies that were working on this before were pipl, spock, yasni, 123people, isearch, wink, etc.


True.. but they all use public records. Use one of these services and find all your friends and old classmates that live in New York. And their UI makes you start by typing a name... we reimagined and redesigned what people search could be.


I for one don't like this, and hope that there is an opt-out feature.

Just because my information is somewhat public in several places (not Facebook) does not mean I want it aggregated and data mined for commercial and advertising purposes.


It's my opinion that once you put information on the internet it is public domain and anyone who wishes to mine it is free to do so. The burden of keeping yourself private lies with you.


I agree with you, the problem here is that being connected on G+, facebook or linkedin to people going on Ark, now Ark will have a lot of info on me that was only available partly to each website. This kind of bothers me.


That is your opinion.

What about companies that resell your data and publish it without your knowledge through a convoluted series of T&Cs?


Very interesting pivot. They were doing an affiliate sales system where rewards from your purchases go to nonprofits of your choosing. They were my startup's #1 competitor. I wonder why the pivot?


I'm detecting at least a small bit of influence from Facebook in the UI.


Pretty fancy detective work there. :) Our beta users felt more "confident" with this design, data layout, probably because they were used to a similar mental model with FB.


That's interesting. I don't necessarily think there's anything wrong with it. Facebook says that they change the font of their site to match the system font of the current user because they want it to "feel like a system application". I guess the analogy is for Ark to "feel like a social network application" by having similar styling. IDK just thinking out loud.


For anyone interested, the closest thing to this around at the moment (that I have come across) is http://pipl.com , and it's actually pretty good.


I agree.. but again, it starts with a "name" input field. Many of the interesting searches are where you know qualities of people, but no idea about their name. It's x10 more powerful than pipl.


I see they're indexing renren– any chance for mixi indexing as well?


Indexing mixi makes a lot more sense than renren since there's no way a site that indexes facebook and google plus will be viewable in China.


Great point! Just as Google first thought they were badass with "1 million pages" indexed, we realize we are just getting started. We'll add more social networks in the future. :)


It is about time someone tackles the problem of so many closed disperate people data sets and find a way to merge them!


What if Google and Facebook block access to their APIs? This seems like a possibility if Ark gains traction.


That's a possibility, but the article states that Ark used its Y Combinator connections to get full approval from Facebook to dredge its Graph API.


They can use the Common Crawl, LinkedIn's API, Twitter's or other APIs. Any single data supplier withdrawal won't disrupt their business model.


If it gained traction, wouldn't a Google or Facebook acquisition be even more of a possibility?


and which ever one loses the acquisition bid will block the winner


That's the winner's problem.


Congrats to Patrick and Yiming on the launch. It looks awesome and adresses real problem.


I want to know what Ark is doing in terms of privacy and opt-out?


Patrick and team are amazing. All the best with Ark and penguins.


This is godsend for recruiters.


Wow, such a great idea, and its presentation video is phenomenal. I don't see how this could not succeed.


I wish there was an API where I could input as much info as I had array("firstName"=>"John","lastName"=>"Doe","zip"=>"12345") for example and it returns info on as many matches that go with the input and includes occupation, birth date, relatives, etc. That would be awesome...


FullContact does something similiar. http://developer.fullcontact.com/


No occupation or it would be pretty good, but that's mainly what I want. Does LinkedIn have an API that this would work with?


just code it yourself, as far as im concerned the website IS the api :)


which site to scrape for occupation - linkedin I guess would work to get decent matches.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: