The official Speechly Quick Start that helps you get started with developing with Speechly.
This Quick Start will guide you through the basics of building Spoken Language Understanding models with Speechly. It covers the following steps:
A good way to start developing with Speechly is to complete this Quick Start and move on to the Web Client Quick Start.
Here’s a short 3 minute video where our CTO Hannes walks you through the setup and initial configuration:
Video coming soon!
The first step is to create an account to Speechly Dashboard and to accept the terms and conditions.
After you’ve created an account, you land on the Speechly Dashboard main page, where you can manage your applications and find example code snippets for building your own client.
Proceed to creating a new application.
When creating a new application, you need to specify a name for your application and choose the language for the speech recognition models. You can initialize the new application by selecting one of our example SLU rule configurations, or start from an empty set of rules.
Custom acoustic models
Now that you have created your first application, you can start adding and editing the rule configurations. This is the most important part of building a well-working voice user interface, so please go and see our more in-depth documentation on Editing NLU examples.
The SLU rule definition works by defining example utterances which have been annotated to specify the users’ intent and entities for the specific utterance. A rule starts with the defining the intent, and then all the entities are annotated into the utterance. As an example, in the rule below “*change” defines that the intent of the specific utterance is change, and two entities have been annotated, wall is an entity of the type object and red is an entity of the type color.
*change paint the [wall](object) [red](color)
With even a simple model you should write several annotated sample utterances in the SLU rules. For more information on writing SLU rules read Editing SLU rules
After you are happy with your new SLU rule configuration, you can proceed in publishing your application and testing the configuration. After you click Deploy`, the new rules are used to train the SLU models and the newly trained model is deployed behind your app-id in the Speechly API.
If there are errors in your configuration, you need to resolve these before being able to proceed with publishing your application.
When your application has been published, the ‘Try’ button below the SLU rule configuration pane becomes active. Use this button to open the Speechly API Browser, where you can test your new SLU configuration in practice.
In the Speechly Playground, you can push and hold the microphone button to speak and see the results for your model.
The results show how the SLU model translated your spoken utterance into text and which intents and entities it identified. While you speak, you will see in the results appearing in the Speehcly Playground in real-time. This should give you some ideas of how to leverage the real-time SLU results in the UI of the client you are building on top of the SLU application.
The last step is to connect the SLU to your own client. To do this you can use one of our client libraries. Happy developing!
Voilà, and you’re done with the quick start! Now you can proceed with the browser client library quick start.
Last updated by ottomatias on March 25, 2020 at 16:46 +0200
Found an error on our documentation? Please file an issue or make a pull request