Create new instance

03 Apr 2014 instance, feature, new

If you are playing with Wit.ai, you may be using your instance like a sandbox to try intents and entities, or test the integration with your app. Once you are ready to start a real project (something new or just the same thing but cleaner!), you can create a new instance.

To create a new instance, click on your GitHub ID on the top right corner of the console and select “Create new instance”.

image alt text

image alt text

And voila! Once you have several instances, you can switch instances by clicking on your Github ID.

You can create as many open instances as you wish for free. As for the number of private instances, it depends on your account type. You can check this via the My account menu.

image alt text

Team Wit


Towards Community

31 Mar 2014 vision, feature, community

It’s been a great start so far! But if we truly want to make Natural Language simple and accessible for the world, we have to join forces. We need a sharing community that can collaborate!

Until now, each developer had their own Wit instance(s). Moving forward, we will start promoting collaboration between developers: you’ll be able to explore and contribute to other instances, as well as get contributions from other developers to your own instance. Just think public GitHub repositories.

image alt text

We’ll start rolling out these features gradually. The first step is the new design of the Wit Console that we released 2 weeks ago. Then on April 8th, you’ll be able to explore other developers public instances, and your instances will become explorable as well.

It is possible to keep your instances private (think private GitHub repos). To do so, you can subscribe to a private plan on your account page.

image alt text

Want to change the world? Sign up here!

Team Wit


Minority Report vs. Her

24 Feb 2014 vision

Have you seen Her? Interestingly enough, the genre of this movie is officially science fiction romantic comedy ‐ a clever attempt to address the problem of couples trying to find something to watch together? but that’s out of the scope of this blog.

The story takes place in 2025. The main UI between humans and computers has become voice. People wear a simple earbud and use it through the day to work, communicate, or play:

Compare with the gesture-heavy UI for which another movie, Minority Report, is famous:

Wow! How many calories spent to put a file in a folder? Very impressive UI, quite visual, that’s good for a movie because it’s spectacular, but is it good for an actual interface you’ll use all day? Here is an excellent in-depth analysis by Christian Brown.

Here at Wit, we prefer the Her approach. It’s seamless and easy, natural. Screens still exist but they are really second-class citizens: for instance in the movie Theodore processes all his messages using just his earbud ; when one of the messages contains photos, he grabs a small screen from his pocket, the pictures are already displayed, he checked them out and put the screen back in his pocket.

Some hardware people are already working on Her-like earbuds: checkout the uber-successful Kickstarter project The Dash.

At Wit, we do believe in this vision too. While some aspects of the AI in Her are clearly impossible to build for the moment, we think that the voice interface is within reach of a community-powered project like Wit.

Interested to build your own voice interface? Sign up here!

Team Wit


Toward Dialog: Contextual Intents

20 Feb 2014 feature, context

Imagine the following dialog:

  • User: Tell Jules I’ll be 5 minutes late
  • Machine: No problem. Do you prefer email or SMS?
  • User: SMS

In order to process this last user utterance, you want Wit to be in a specific state where it expects things like SMS or email.

Following last week’s announcement, many of you asked us how to manage this in Wit. Well, actually it was not possible! But now, it is. Here is a short screencast that demonstrates this new feature:

So now, some intents can be activated only in certain states:

  1. Navigate to the intent page

  2. Click the “Add State” button, type a state name, then Enter.

image alt text

When you do an API call, use the optional context field. It should be a JSON-encoded object:

{
    "state": "email_or_sms"
}

Check out the full documentation about states.

Team Wit


Introducing Wit Speech API

12 Feb 2014 feature, speech

Voice commands are the future. Science-fiction has had them for decades and yet, we still have reach for the remote to turn on the TV or set an alarm. Our mission is to change this. Adding a voice interface to an app or device should be simple.

Turning speech into actionable data

Today, we’re very excited to announce our new “Speech to JSON” API, four months after the launch of the “Text to JSON” API.

speech_to_json

From now on, your app, device or even your website can stream audio to our server, and get actionable data in return.

See it in action for home automation:

How does it work?

Behind the scene, Wit combines various state-of-the-art Natural Language Processing techniques and several speech recognition engines in order to achieve low latency and high robustness to both surrounding noise and paraphrastic variations (there are millions of ways to say the same thing).

Fortunately, you don’t need to care about all this machinery. We focus all our energy into creating the simplest developer experience possible. You can be up and running in a few minutes using our website. Wit will adapt to your domain over time, from ice-cream distribution to space missions. Wit makes no assumptions and remains 100% configurable.

It will take you 5 minutes to build your own Wit configuration:

Consuming the API

Then, calling the API is simple. We provide client-side SDKs that handle audio recording and streaming for iOS, Android or even a simple webpage like this one. You can also use the HTTP interface to stream live audio or post a sound file:

Let’s take this sound (recorded from a celebrity in the valley – do you know who?):

Submit it to the Wit API with a POST request:

curl -XPOST 'https://api.wit.ai/speech' \
     -i -L \
     -H "Authorization: Bearer $TOKEN" \
     -H "Content-Type: audio/wav" \
     --data-binary "@sample.wav"

You’ll get this in return:

{
  "msg_id" : "6a84eae3-969c-41ad-94d9-85076fbbdc99",
  "msg_body" : "set the kitchen table on fire",
  "outcome" : {
    "intent" : "set_fire",
    "entities" : {
      "object" : {
        "value" : "kitchen table",
        "body" : "kitchen table"
      }
    },
    "confidence" : 0.997
  }
}

Interested to build your own voice interface? Sign up here!

Team Wit

@WitNL