Develop Speech Recognition Enabled Apps with Api.Ai

“What’s on your mind?” Facebook asks me every time I log in. A lot, I shout!! Sometimes so much that I am not able to articulate it into words and type elaborate posts! I wish it could really read my mind and pen down my feelings for me…I wish there was a robot who could multi-task for me and help me in keeping my sanity.

When I tell my husband this he says the day is not far away when not only Facebook but many other applications, websites, and devices will be at your voice command! Siri (Apple), Cortana(Microsoft) and Google Now (Android) are already making people’s jobs easier.

This conversation happened yesterday and today you can imagine my surprise when I heard about Api. Api is a software that enables you to create Speech-Recognition enabled Apps!

The more I looked into it as an App developer the more it intrigued, excited me! I have developed an App called ‘Ask Me Anything’ which guides people on diverse subjects. I have a team of experts from different fields like law, accountancy, and real estate, technology who tries to answer people’s queries as best as they can and I get more than 200 queries a day!

Imagine if people didn’t have to type those queries? Just open my App and talk! Once the Api has been implemented, it will give my App the ability to listen to voice commands, accept them and carry out the given command.

Here is what I found out about Api.Ai :

  • It is easy to implement and your voice-enabled App can be developed in a matter of few hours!
  • Firstly one will need to identify and describe the objects that will interact and work with the App.
  • Secondly, you’ll need to make the actions or scenarios that should be performed. E.g. what the App user may say based on what your App is about…say in case of my App the user will ask “Can I have a list of apartments for sale on Michigan Avenue?”

This will probably be the tedious process as you will have to imagine all the possible things the App user may say but it will be a worthy exercise to give your App the required edge over others!

  • Finally a little bit of integration work and your App is voice enabled!

 

 

Here is why I would go for Api.Ai without a second thought:

  • It gives me a complete Speech Interface with
  • speech recognition,
  • intent recognition,
  • context management,
  • conversation management and
  • command fulfilment
  • I know for a fact that many App users skip the instructions, tutorials, get flustered with the options and features and then give bad reviews about the App. I want to make their work easier and mine too!
  • It has Knowledge bases for multiple domains which is an added advantage!
  • It offers ‘Learn by Examples’ service whereby if we give them a few examples, they will do the rest of the work!which means more time for me to focus on upgrades, additions, fixing bugs than on mundane work!
  • Its multilingual! As of now offering 11 languages support with more to be added soon. Gives my App a worldwide reach.
  • It offers support not only for iOS but for Android, Microsoft too.
  • It heralds endless opportunities not limited to smart phones but it can used for websites, PCs, Televisions, Music systems, Car Systems etc
  • It is a high performance platform and has been put to test by way of more than billion requests!
  • A voice enabled App is bound to be more attractive to the user and more likely to be downloaded and used.

 

Here is How Much it Would Cost Me:

There are different packages available and the pricing seemed very reasonable to me.

  • For non-commercial, free and open source projects the charge is a big ZERO! It’s free of cost with full access to all its tools!
  • For commercial projects there are two options:
  1. You can go for revenue sharing program or
  2. Pay a specific amount per request, starting from as low as $0.003.

If usage of the api.ai by your application is projected to consistently exceed the monthly usage limits, you should enable billing on your api.ai project for additional usage.

Once billing is enabled and your application has exceeded the api.ai free usage limits, excess users over the usage limits will be calculated daily for your application, and billed to your credit card at the end of each month. Your bill will be prorated until you reach the next tier.

Here are some of the hiccups I expect along the way:

  • Accent recognition.
  • Users saying non generic words or slang.
  • Lack of database in a particular field.
  • Changes in the quality of API over a period of time.
  • Higher average API response time and higher error rate.

But these seem to be minor compared to the benefits and can be overcome easily.

Here is Why I Believe That Api.Ai is ‘The Future’!

The other day my kid could not find the TV remote and was very upset. I told him that there was a time when TV had no remote and you had to walk over to reduce the volume and change the channel.

Saying that didn’t help at all, but I got wondering that how we have become so technology driven! As an App developer you need to future-proof your cross platform development, keep abreast with the sea of changes and offer the best to your consumers!

Offering speech recognition enabled Apps to the consumers through Api.Ai will be great armour in the competitive battleground of Internet.

Offer comfort, offer convenience to your customers – Imagine walking into the house and commanding – Lock the garage door, preheat the oven to 250, open the blinds, turn on the Wimbledon match between Federer & Djokovic, call your favourite Chinese take-out restaurant! All in a matter of few seconds! Bliss indeed!

Leave a Reply

Your email address will not be published. Required fields are marked *