Chatbots made easy

ChatAlly is a collection of modules that help you build your own self-hosted chatbots. If you know, how to build websites you won’t have much trouble to build chatbots with ChatAlly. It is implemented in JavaScript, so you do not have to learn another language when already learning a new technology. But we also support integrating solutions written in Python or other languages.

Integration

ChatAlly is all about integration: Use your machine learning framework of choice to create a natural chat experience, e.g. Huggingface’s transformers.js, TensorFlow.js or other more dedicated chat frameworks like nlp.js.

Documentation

ChatAlly gives you all the tools to get started quickly and evolve your chatbot into your very own support character for your users. To support you on your journey, documentation is essential, that is why it is an explicit work package in our project and will account for at least half of the effort. Check out our guides and examples.

Servers

Connecting your chatbot to the channels that your users use is often a challenging task in the beginning, because of the proprietary protocols being used. That’s why we offer reference implementation of servers, talking to your WhatsApp, Signal or web channels.

Middleware

Easily integrate all kinds of middleware to create a rich user experience. Natural language processing is only one part, add a consent manager, log training data from real chats, integrate your knowledge base or speech features. If you can’t find, what you need, write your own middleware easily.

Community

Share your own middlware with others, get help from experienced users, share ideas for new features or fix a bug. Checkout our Github project or join our Discord server (soon to come).

Who is behind ChatAlly

For now it is only me, Christian, co-founder of the non-profit organization dooiy, that provides step-by-step instructions for good hacks, that make life easier. dooiy’s focus audience are users in underserved communities, i.e. mainly people in slums or rural areas living in extreme poverty.

In this context, for a few reasons, we were looking for an easy way to distribute these instructions via WhatsApp. Only to find out, that this becomes very expensive or complicated, if you want to do it right and according to European data protection laws. Because of lack of money, we chose the difficult route and implemented our own solution that allows us, to (fairly) easily create chatbots from Markdown files. It is currently being prototyped in Mali with an instruction from CoolVeg for a clay pot cooler.

I proposed to the Prototype Fund to enhance this solution and make it free open-source. This is how ChatAlly was born. I am currently in the process of developing the first version of ChatAlly and am looking for contributors and users to help make it a success, so that it can empower small organizations and communities to deploy their own chatbots affordably and GDPR-compliant.

While there may seem to be open-source or free platforms at first glance, the quotas of free offerings are severely limited, and open-source typically only refers to the Software Development Kit (SDK). Data control usually lies with the platform providers. True FOSS alternatives such as frameworks and libraries for chatbots are complex to set up and do not support seamless integration with popular messenger clients.

With ChatAlly, our aim is to bring these components together into a plug-and-play solution that can be easily self-hosted.

Impressum