Return to site

Four problems facing the API economy

A number of significant developments over the past 5 years have propelled us into what some call an API economy. APIGEE has a great video explaining this which is definitely worth seeing. In a nutshell, because of the rapid proliferation of devices and apps, if you have some data that is remotely useful, it needs to be accessible through an API. This has led many companies – big and small – to develop proprietary APIs that are either public or accessible to partners.

While API technology has also matured over the past 5-10 years, developing applications against an API (or a set of APIs) is still not easy. Some of the more established APIs are not necessarily the most modern in design, which makes integration a challenge requiring time and valuable resources. Often these happen to be important APIs that many people have to integrate with.

So there’re two main problems here: First, there are too many APIs, and second, many are not easy to integrate with. There’s actually a third – hidden but significant – issue in that there isn’t enough uniformity in the syntax and semantics of APIs even within a single industry. While REST has done a lot to improve this situation, there’re no “enforced” rules governing syntax and semantics.

Fortunately, we believe there is an elegant solution that addresses these problems, and we call it API Aggregation. The idea is to bring key APIs within an industry under a single umbrella. It would act as a gateway to the data available through those APIs by providing a single uniform and modern API. The ease and speed of integration that this approach provides are significant. Improvement in time-to-market and access to the pool of untapped data can open up many possibilities.

API aggregation actually solves a fourth problem that can be a show-stopper in any API integration project. APIs change all the time and keeping up with the change is not easy. By shifting the burden of upkeep to a single provider, developers can rest assured that their work doesn’t break and they can stay focused on data analysis and visualization, rather than data collection and integration.

(Note: This post was originally published in December 2013.)

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!