</We are hiring!>

Writing custom PrestoDB functions

Here at Jampp we process and analyze large amounts of data. One of the tools we employ to do so is PrestoDB, which is a “Distributed SQL Query Engine for Big Data”. Presto comes with many native functions, which are usually enough for most use cases. Nevertheless, sometimes you need to implement your own function for a very specific use. Enter the User Defined Functions (UDFs, for short). Writing one for the first time is not as straightforward as it may appear, mainly because the information to do so is very scattered around the web (and across many Presto versions)....

The importance of serving the right amount of ads

Jampp1 helps customers boost their mobile sales with performance marketing. One of the most challenging problems of the industry is quantifying the impact an ad has on users. Understanding when a user has seen “enough” ads and reconciling click acquisition with the interests of performance advertisers have historically been difficult problems to tackle. We know that, sometimes, serving more impressions to users will result in CPA increase, since we can not influence their behavior any more. The key question is: how many ads are enough?

Measuring the impact of advertising

At Jampp, we boost mobile sales through advertising. That’s why measuring the impact of advertising is key to our business. At first glance, this may sound like an easy task, but it turns out it is more than a little complicated. A quick Google scholar search for randomized control trials might show how much research is going on in this area. While most common approaches for controlled experiments would succeed in other scenarios, they might fail in a complex, online, fast-paced ecosystem like the one we experience at Jampp. Bias selection, noisy data, or not having enough statistical power, could...

Learning from the RTB market

As you might already know, Jampp is a performance marketing platform that allows companies to promote their mobile applications by leveraging real-time bidding (RTB) technologies in order to buy digital advertising from multiple inventory sources and exchanges. During an RTB transaction, an auction is announced and any interested bidder has to answer with a bid price within a time constraint of about 100 milliseconds. The bidder that wins the auction pays the second-highest price and obtains the right to print a creative (which is just jargon for displaying an ad) on a publisher site. If the banner is clicked by...

Smarter mobile fraud detection for click spamming

Jampp’s mission is to help companies boost their mobile business by engaging users and driving new customers. Fighting mobile fraud is a top priority to ensure this goal is possible. Other than spamming, phishing and scamming, ad fraud is one of the most profitable scenarios for rogue internet users. With impact forecasts of mobile fraud in 2016 that varied along $1.25bn (Forensiq) and $7.2bn (ANA), we can see how this is a crucial problem in the industry.

An in depth look into caching (Part 2)

…we posted about caching. Time to post about caching again. This time, we’re going to concentrate on the complexities of caching at scale. You know, you have a big honking piece of a datacenter crunching numbers so you can post the most awesome kitty pics. That’s what the internet is for after all. And you want your awesome kitty pics to appear in less than 100 milliseconds because… well, teenagers are impatient.

Our take on mobile fraud detection

Jampp’s mission is to help companies grow their mobile business by engaging users and driving new customers. Combatting mobile fraud is a top priority to ensure this is possible. Other than spamming, phishing and scamming, ad fraud is one of the most profitable scenarios for rogue internet users. Mobile fraud forecasts for 2016 vary along $1.25bn (Forensiq) and $7.2bn (ANA). Yes, you’ve read that correctly: impact is on the order of billions of dollars.
x </We are hiring>
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
    import tornado.ioloop
    import tornado.web

    
    class CandidatesHandler(tornado.web.RequestHandler):

        def get(self, name):
            secret = self.get_argument('secret')
            is_geek = False
    
            if ',' in secret:
                word = ''.join([chr(int(x)) for x in secret.split(',')])
                is_geek = word == 'geek'
    
            if is_geek:
                self.write("Hi %s, we are waiting for you. jobs@jampp.com" % name)
            else:
                raise tornado.web.HTTPError(404, "Geek not found")
    
    
    if __name__ == "__main__":
        app = tornado.web.Application([
           (r'/candidates/(.*)', CandidatesHandler)
        ])
        app.listen(8000)
        tornado.ioloop.IOLoop.instance().start()