Sounding like AI and being able to actually hold a conversation while in character are two different things though. Some really fancy ones are getting really close. Source: over 1 year ago

For example: https://ai.googleblog.com/2018/05/duplex-ai-system-for-natural-conversation.html. Source: almost 3 years ago

I'm working on a game that has an AI narrator talking to the player. I messed around with some filters in Audacity that I don't know how to use until I realized I don't want a stereotypical computer voice, what I'm actually looking for is similar to Google's creepy Duplex system that automates calling restaurants for reservations. Source: about 3 years ago

Beyond raw language models, NLP engines like Rasa and Dialogflow offer frameworks for designing, building, and improving conversational flows. They help in intent recognition, entity extraction, and dialogue management, which are crucial for a coherent conversation structure. - Source: dev.to / 3 months ago

There are frameworks out there for doing that kind of thing, see https://rasa.com/ for example. It's not using any LLMs at the moment, just BERT and DIET mostly but it's highly customizable and you could likely bring in an LLM for doing some interesting things to handle more complex messages from users. - Source: Hacker News / about 1 year ago

Chatbot frameworks: Utilize chatbot frameworks such as Botpress, Rasa, or Microsoft Bot Framework to streamline development. - Source: dev.to / about 1 year ago

Rasa is a popular tool used right now to build these applications. If you're looking for a serious turn-key solution I would check out Vectara. Source: over 1 year ago

Another example is RASA, one of the most popular platforms for creating conversational AI assistants. On the accepted AI quality scale, RASA reaches levels 3 and 4. It means that the "robot" not only understands humans with high accuracy in a given contextual field but also learns to recognize contradictions and ulterior motives. - Source: dev.to / over 1 year ago