Roll your own chatbot with Rasa
Causal Chat

Many websites offer an automated feature that responds to visitor questions. If you're ready to build your own web chatbot, but don't have time to become an AI expert, Rasa Open Source can help.
You can imagine a chatbot as a kind of robot that talks to you. Underpinning the chatbot experience is software that can analyze language in text form and find an appropriate response. The technical term for this process is Natural Language Processing (NLP). In the past, NLP systems were completely rule-based. To be more precise, state machines were used to conduct the conversations.
However, rule-based chatbots were extremely limited. They were only suitable for processing simple standard tasks. A chatbot could provide information about a specific item in an online store or the opening hours of a museum. However, recent AI techniques allow for a much greater range of expression. Modern AI-based chatbots are very capable of conducting complex conversations.
Modern chatbots start by analyzing the request and trying to understand what it means. This step is known as Natural Language Understanding (NLU). The result is machine-readable information, such as keywords or specific email addresses.
[...]
Buy this article as PDF
(incl. VAT)