Today in APIs: Breathometer Connects Drunks to Uber with API

Breathometer, the maker of a bluetooth connected breathalyzer, uses Uber's API to connect people who need a ride. Newsly, an overnight hack, provides tailored news using Machine Learning APIs. Plus: hackathons head to rural colleges, and Infected Flight is the timely winner at Disrupt Europe hackathon.

The Breeze Breathalyzer API Steers Tipsy Users to Uber

Breathometer has released its improved Breathalyzer, Breeze, which offers law-enforcement grade accuracy. It offers users the chance to grab a ride with Uber, and it's looking to expand those offerings, to hotels, for example, for those wishing to sleep it off.

But as Ryan Lawler at Techcrunch reports, the company aims to do more than stumble around in the market for alcohol testing:

In addition to being able to detect blood alcohol content, the sensor can be used to take other readings. For instance, it could be used to detect halitosis or dehydration, according to founder and CEO Charles Michael Yim. And that’s all part of Breathometer’s bigger play — to provide a series of devices that can be used for different applications, all with a similar design and tech embedded.

In addition to future devices, the company seems bent on squeezing every use it can think of out of this one. The iOS version of the app connects to Apple's Healthkit, allowing users to input data on their blood-alcohol content levels into their health statistics.

Newsly Powered by Machine Learning APIs

Newsly, imitating Tinder's popular method of liking and discarding potential dates by swiping right or left, offers users the chance to like and dislike news. Using machine learning APIs, Newsly then crafts a profile to generate tailored news. This has a leg up on news feeds from the likes of Facebook, which generates feeds based only on likes. As Natasha Lomas explains in Techcrunch, the revenue stream is obvious: advertisers will pay for the preference data. She explains how it works, quoting one person on the three member team, Muhammad Rafdi:

For tailoring news content to personal preferences the app is doing natural language processes. “It searches through the [news] content — it does a text analysis, and the text analysis will produce keywords which are then stored into the server,” explains Rafdi. “After several queries it Builds up a new one and then that will Fetch a new list of news Feed based on your preferences.”

This power of dislike has broad implications, so much so that we can imagine Facebook eventually embracing it. By using only "Like," Facebook is giving people--and advertisers--one half of a binary choice. Facebook's problem, of course, is that unleashing a dislike button could have all kinds of nasty consequences when people, news, and products get dissed. But what makes Newsly work is that those preferences aren't shared publicly: the news you have no use for does not get branded as dislikable. What's to prevent Facebook from doing likewise and offering a dislike button that builds data on preferences without pasting the results on the content provider, user or company's pages? The app uses the Intel Mashery API to pull news from the Guardian, though it will soon be used to pull news from many different sources.

API News You Shouldn't Miss

Be sure to read the next Safety article: BlueLight Chooses Bandwidth API Platform for Added Campus Safety