Hu:toma Open Source Conversational AI Platform
Hello! We are thrilled that you are considering our platform for your conversational AI projects! Below few pointers that we hope will get you started. For anything else, please join our community
Hu:toma is a fully functional and scalable Conversational AI Platform that was designed to simplify the process of creating and managing chatbot projects. You have different ways to access the platform:
- Online Demo: An online demo is available at this address: All the functionalities you see there are available as open source
- Docker Image: You can deploy the entire platform locally or in your private cloud as container. Instructions on how to do that can be found on github
- Source Code: Source code for your own build is published in our git repository
Here's also a brief orientation video and some common terminology that you might find useful if you are new to the Chatbot world.
A bot or chatbot is an Artificial Intelligence machine that responds to user input, as entered in normal, “natural" language. The bot's response is also in natural language so that the user has a "conversation” with the bot.
User-to-bot conversations can be on any topic that the bot developer designs. The conversation can also provide a service, such as looking up the weather, providing customer support, or booking a cab. This is entirely for the bot developer to decide.
A client is a way for an end-user to access your bot. This could be hosted on your website, or integrated to some third party service such as Facebook Messenger or Slack.
The developer console is a website that a registered user of our platform goes to if they want to:
- Create a bot
- Edit a bot
- Test a bot
- Manage the bot lifecycle
An entity is a category of items like Cities in the USA or Coffee types available on my menu, along with a list of valid values for that category. One or more entities are required by an intent.
An intent is the capability to capture context data from your chat users that allows you to take an action, either in your client or using a webhook.
An intent requires one or more entities to be matched.
For example, a weather bot would have an intent with a location entity for where to get the weather for. Then a webhook for the weather data REST API would be called, passing in the recognized location.
A REST API provided by a service such as a weather data provider or a restaurant chain's booking. This can be used to do a real-world action when an intent is matched.