Build your custom AI chatbot with RAG

You can feed your chatbot with various types of data: plain text, PDF documents, images, audio, etc. You can also provide links to web pages: we've integrated a webscraper that extracts the text itself.

No! You can either add it manually from our interface or via API. So you can add new data as you go along, enriching your chatbot with new knowledge.
‍

Yes, we've done everything we can to make creating your personalized chatbot as simple as possible. You can either customize the parameters for creating your chatbot (chunk type, vector database, choice of AI model, etc.) or use the configuration we've set as default. Our support team is also available to help you if you need it.

Costs will depend on the AI models you use to run your chatbot. There are at least two models: the embedding model, which will index your data, and the LLM, which will query the data and generate the response text. If you're indexing non-textual data (PDF, audio, etc.), other models may come into play (OCR, speech-to-text, etc.). You can easily track these costs on our monitoring tool.

You can use the chatbot to integrate it into various applications. For example, you can use it in messaging systems such as Discord, Slack, Intercom, WhatsApp and so on. All you need to do is request your specific custom chatbot API.

You can also embed the chatbot on a website or application. We provide a piece of code to be integrated into an HTML code that displays a chatbot for your users.