I recently designed and built a web app that depends on a fair-sized database and a broad collection of queries. I didn't write a single line of API code to implement the backend data integration with the application.
Beyond APIs there exist a new world of real-time data driven by web sockets. Sockets-based architectures allow us move data far quicker and easier than through traditional API integrations.
How It Works
Imagine when a web page is opened in a browser, a team of minions instantly lay a big pipe between the web page and a database. When the pipe is created, the web page drops into the pipe, a simple message -
Send me everything you have concerning (x) data.
And instantly, the data flows from the pipe in the form of JSON objects. Here is how it's done in the web app.
The web app has all the data about (x) and is ready to use it as needed. If new records are added to the database from other sources or other web clients, the web page is instantly updated with the new additions - in real-time. If you also instrument a pipe to feed the database, the updates are instantaneous as if you had chatted the data over to the database.
This architecture is virtually identical to your texting app on your smart phone. You type --- your spouse sees it instantly; she types -- you cannot deny that you got her message because you almost certainly did get it ... instantly. ;-)
The Real-Time Economy
Just as texting has become a universal API for human-to-human communication, data interchange through sockets will take us far beyond the constraints and latency of traditional APIs.
Indeed, we've only recently embraced the API economy and upon us today is the Real-Time Economy; an emerging technology trend that will virtually eliminate latency while transforming the way that we move information around the interwebs.
Humans are selfish; we want information now, and in some cases, we must have it instantly because it has a shelf life that is shrinking in direct relationship to our attention span.
Real-Time Use Cases
In security situations, latency is trouble. Time-to-awareness is critical; the shorter the better. And instantaneous is as fast as it can get.
Business analytics is another horizon of opportunities exploding with demand.
Imagine you're the manager of a KFC. Real-time analytics would allow you to know exactly how many people are in the queue at the front counter as well as the drive-thru. You could also be aware of the make, model and number of people in each car.
Imagine you are sitting in front of a map. It's the floor plan of a 250,000 square foot casino. A known thief strolls in a back entrance and a Canon M50B recognizes this person's face. The IBM software is able to perform computations to identify the individual and through a traditional API, it pops up a message and pin marker on the map while zooming to that location. This takes 30 to 60 seconds.
What if the camera, the recognition software, and the map were all connected through a sockets layer? The performance improvements are astounding. We can reduce the time-to-awareness by 25 to 55 seconds.
The ability to grab some bad guys before something bad happens is improved significantly. The ability to push marketing content more rapidly is realized. The capacity to act with greater precision for security, marketing, and automated events is transformed as we approach zero latency in automated processes.
The Internet of [Instant] Recognition is at the doorstep kicking the screen door. What will you do with this advantage in the Real-Time Economy?