3 API trends shaping the future of apps

Written by Colin Morris
Published on Nov. 17, 2015
3 API trends shaping the future of apps
 
For the uninitiated, the term Application Programming Interface (API) may just sound like more of the alphabet soup sloshing around the tech world. But APIs are critical to a lot of software development, not least because they provide the backend building blocks for importing and manipulating data to be put to work in new ways.
 
Want to build an app that tracks how the US government spends money? You can, because it provides APIs for its public data. Same with third party apps that track public transit arrival times, flight delays, social media feeds and more.
 
But with so much data to track, these apps can only be as good as the APIs behind them. That's why it’s more important than ever that APIs work well.
 
That’s the theme of this year’s API Strategy and Practice conference, where experts will discuss how best to harness data over three days. We asked four of the speakers for the most important roles APIs will play in the future and how best to manage them.
 
Here are their top 3 tips.
 

3. APIs will drive artificial intelligence

For Ashley Hathaway (pictured below), Senior Developer Evangelist at IBM Watson, the future of APIs leans a little more sci-fi. Watson uses natural language processing to consume massive volumes of unstructured data and learn from it, eventually using its expertise to formulate output through cognitive computing.
 
“Thinking ahead to 2016, it’s all about democratization of data, IoT and machine learning,” Hathaway said. “Developing APIs [to parse data as complex as] machine learning, AI, neural nets, neuro-linguistic programming and other algorithms will allow developers to work with data in wholly new ways.”
 
Essentially, the faster we generate data and the more widely available we make it, the more inventions will emerge to take advantage of it.
 
“Through this democratization of data AI APIs will continue to affect our daily lives more and more through things like self-driving vehicles, expert systems and personal assistants,” Hathaway said. “That’s not to say computers are anywhere near ‘taking over.’ [They will] only serve to surface more valuable information for us to make better decisions and ask better questions. By 2017, development teams will have more data scientists and engineers, and we’ll start to see more and more Machine learning as as service APIs take form.”
 

2. APIs will become software products

James Higginbotham (pictured right) is the founder of Austin-based LaunchAny, a tech consulting firm focused on APIs and mobile development.
 
“Businesses that are most effective at managing and using APIs are the ones that productize their APIs,” he said. “They have moved beyond building single-use APIs that solve an immediate problem, such as powering a single mobile application. Whether they are an early stage startup or an established business, they see their APIs as strategic to the business.”
 
Higginbotham said the first step in that direction is to apply systems design and domain driven design techniques to ensure your API has clear boundaries. Then set it up to solve a variety of use cases for the intended audience.
 
Robert Read, co-founder of the White House's API-focused 18F lab, was even more tactical.
 
“You need to treat APIs just like software and manage your release numbers for your APIs. Hopefully the API will not change as often as the underlying implementation, but it will change. You can make it easier for your users if you have a numbering scheme that makes it clear it clear when you are making a breaking change that’s not backwards compatible.”
To that end, software engineer Bonnie Mattson (pictured left) said it helps to have the right tools.
 
"Logging and monitoring tools are essential," she said. "At Context.IO, we use Scalyr as our log analysis tool and a combination of Scalyr, Datadog, Runscope tests and our own integration test suite to monitor our API health and trigger alerts on failure states."
 

1. APIs are a public service

Read advocates better software development as a public service of the federal government, which has hustled under the Obama administration to catch up to modern standards and expectations of transparency through technology.

He draws a bold line between APIs themselves and the Graphical User Interfaces (GUI), or user-facing apps, developers build to leverage the API.
 
“I would also assert that it is the government’s first responsibility to produce an API. I mean that it is even more important to produce an API than a GUI. This smacks of elitism, because not everyone has the knowledge to use an API. However, not everyone has the knowledge to fix their own car or defend themselves in a criminal court, either. What is important is that we give people the ability to use an API even if they have to hire a professional to do so. In the past, governments have not thought this way, and have produced GUIs first and APIs as an afterthought, if at all.”
Hiring Now
Digital Turbine
AdTech • Information Technology • Marketing Tech • Mobile • Software