Shortly after the Zypr voice API came on to the scene in 2011, I launched my research into voice APIs. Like many other areas of the API universe, voice has come in and out of focus for me, something I think will take much longer to unfold, than any of us could have ever imagined. Zypr quickly ran out of steam, and other similar solutions have come and gone over the last couple years as well, leaving my research pretty scattered across many different concepts of how voice and APIs are colliding--lacking any real coherency.
I took a moment last week to take a fresh look at my voice API research, because of a comment by Steven Willmott (@njyx), the CEO of 3Scale. Its not an exact quote, but Steve spoke about how voice is the future of API consumption, after he had attended the AWS:Reinvent in Las Vegas. I agree with him. Voice APIs is a topic that has been significantly stimulated with the introduction of the Amazon Echo platform, but I also feel also coincides with a critical mass of available API driven resources that will deliver some of the value these platforms are promising users.
Voice recognition has always been something that leaves a lot to be desired--think Siri. Even with these challenges there are many dimensions to the voice API discussion, and with the amount of resources now available via simple APIs in 2015, I feel we are reaching a more fertile, and friendly time for voice solutions to return the value end-users desire. We now have a rich playing field of weather, news, stocks, image, video, podcast, and other data, content, rich media, and programmatic resources, which can be linked to specific voice commands--something we didn't have before.
While there is still so much work to be done, but I agree with Steve's vision, that voice will play an increasingly significant role as an API client. I would add that like mobile, or the recent wave of wearables, voice will have special constraints when it comes to API design, further requiring API providers keep their APIs simple, and reflect how users will experience them, not just being a SELECT * FROM table WHERE q = 'search', with a URL bound to it.
I think the API providers who are further along in their journey, will get a boost as voice evolves as an API client, and voice enabled app developers are able to easily integrate valuable API driven resources into their solutions. Even with my new found optimism about voice APIs, I still think we are years away from voice solutions actually living up to, even a small portion of the hype they seem to get over the years. Regardless, I'll be working to keep a closer eye on things, and will be sharing via my voice API research.
