Speechly Cloud is the easiest way of deploying and running a Speechly application. It requires minimal setup, supports instant deploys and offers easy scaling. Just create an application, integrate it to your project and start transcribing audio.
Any time you deploy an application using Speechly Dashboard or Speechly CLI, you deploy the application to Speechly Cloud. If you have followed our getting started guide, you have already deployed your application to Speechly Cloud and used it to transcribe streaming and pre-recorded audio!
The main use case for Speechly cloud is development and testing. The Starter plan includes 50h fo API quota, which is a generous amount that allows you to get started with Speechly. When going into production, you can use Speechly Cloud as your deployment option. However, most of our customers choose Speechly On-device or Speechly On-premise for that.
Creating and deploying applications
Both Speechly Dashboard or Speechly CLI allow you to create, edit and manage your applications.
- Open Create a new application
- Give your application a Name
- Press Create application
- Once created, you can make changes like editing the preferences or adding training data
- Press Deploy to deploy the changes
Create an application:
speechly create "My app"
The command prints out the path to
config.yaml file and the
App ID for the new application. Open the config file to make changes, like editing the preferences or adding training data. See available options.
speechly deploy YOUR-APP-ID /path/to/config
Development and testing
Both Speechly Dashboard and Speechly CLI heavily rely on Speechly Cloud. Both offer various tools for testing your speech recognition applications, for example:
- Preview - Use the Preview tab in Speechly Dashboard to preview changes in your application. It uses the browser microphone and transcribes speech in real-time, outputting any intents and entities specified in the training data.
- Evaluate ASR - Evaluate the ASR accuracy of your application using Speechly CLI.
- Evaluate NLU - Evaluate the NLU accuracy of your application using Speechly CLI.
See your projects API usage in Speechly Dashboard by going to Project settings → API usage or by using the
You can integrate Speechly to your product using our APIs or API clients.
Speechly offers two APIs:
- Streaming API - used for real-time spoken language understanding.
- Batch API - used for asynchronous spoken language understanding.
To make coding against Speechly Streaming API easier, we provide API clients for popular web and mobile platforms. The clients handle authentication, audio capture, network streaming and connection management.
Regardless of which platform you choose, the process is quite similar:
- Install & import the client library
- Create a new client instance, passing the
App IDof your application
- Capture audio form the device microphone and attach it to the client
- React to API updates
Follow the Building a React app using Speechly React Client tutorial to learn more about how to integrate Speechly application to a React project. Also, check out the Speechly React Client example application, it showcases the usage of Speechly React Client.