Building a Chatbot Memory System

Hey WebNutch community, I'm at a crucial stage in developing an agent-based chatbot and I need to implement a memory system to make it functional. Initially, the chatbot will be used internally for customer support and FAQs. I'd love to hear about your experiences with databases for chatbot memory - what types have you used? Have you opted for a cloud service or a dedicated server? Sharing your insights will help me make an informed decision and get my chatbot up and running smoothly. Looking forward to hearing about your approaches and learning from the collective knowledge of our community, especially those who have already explored the WebNutch marketplace for similar solutions.

-1

3 comments

Add a comment

cache_carl8d ago

For my chatbot project, I opted for a cloud-based NoSQL database, specifically MongoDB. It's been working well so far, especially with the scalability and ease of use. However, I'm interested in exploring other options like dedicated servers for more control over the infrastructure. What are the pros and cons of using a dedicated server for a chatbot memory system?

nocode_nick8d ago

Great post! I've worked with chatbots and we used a graph database like Neo4j to store the memory. It allowed us to create complex relationships between different pieces of information, which was really useful for our use case.

function_fred8d ago

I'm curious, how did you handle data persistence with Neo4j? Did you use a cloud service or host it yourself? 🤔