Ark (YC W12) launches a modern people search engine

http://techcrunch.com/2012/03/25/ark-people-search/

by priley • 14 years ago

132 53 14 years ago

K2h• 14 years ago

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

mindstab2• 14 years ago

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

tptacek• 14 years ago

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

jaredsohn• 14 years ago

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

priley11• 14 years ago

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.

priley11• 14 years ago

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. :)

mindstab2• 14 years ago

:) 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.

cantbecool• 14 years ago

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

epikur• 14 years ago

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

moylan• 14 years ago

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

chrisaycock• 14 years ago

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.

jordhy• 14 years ago

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.

dm8• 14 years ago

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

suking3• 14 years ago

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...

sumukh1• 14 years ago

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

suking3• 14 years ago

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?

kami8845• 14 years ago

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

suking3• 14 years ago

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

kmfrk• 14 years ago

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

evoxed• 14 years ago

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

dsrguru• 14 years ago

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.

priley11• 14 years ago

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. :)

ayanb• 14 years ago

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

btucker• 14 years ago

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

omarish• 14 years ago

priley must have amazing negotiation skills.

priley11• 14 years ago

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.

bennyfreshness• 14 years ago

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?

sciurus• 14 years ago

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

prawn• 14 years ago

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

useflyer• 14 years ago

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

priley11• 14 years ago

:) Ha.

priley11• 14 years ago

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.

joejohnson• 14 years ago

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.

citricsquid• 14 years ago

indeed: http://i.imgur.com/67KXM.png

priley11• 14 years ago

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.

piggity• 14 years ago

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.

vasco• 14 years ago

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.

koopajah• 14 years ago

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.

newman3142• 14 years ago

That is your opinion.

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

goronbjorn2• 14 years ago

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

priley11• 14 years ago

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.

goronbjorn2• 14 years ago

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.

evoloution• 14 years ago

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...

priley11• 14 years ago

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.

Herald_MJ• 14 years ago

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.

priley11• 14 years ago

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.

stfu2• 14 years ago

"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.

prayag• 14 years ago

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

alagu• 14 years ago

This is godsend for recruiters.

newman3142• 14 years ago

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

Ark (YC W12) launches a modern people search engine

Related Stories