How can i make a AI chatbot with Amazon Bedrock that use my website data for answer the questions?

Question

I currently have a website that has an FAQ section, a tutorial section, and a guide section. I would like to use this information as a basis for the bot to answer the questions from the users.

I know there is a way to do this using a knowledge base, but it requires somehow processing my site and uploading the content to S3, which I would like to avoid.

Looking into the Amazon Bedrock workshops, there are ways such as including all the context in the question, but I'm not sure if this solution could increase too much the costs.

Do you have any solutions tutorials or advice for implementing this?

Thank you very much.

Answer

Another approach is to use Opensearch serverless and Bedrock (without Knowledge Base) : https://github.com/aws-samples/rag-using-langchain-amazon-bedrock-and-opensearch

Answer

Hello.

I think it is possible to use a web crawler by configuring RAG by combining Amazon Kendra and Bedrock.  
https://aws.amazon.com/jp/blogs/machine-learning/index-your-web-crawled-content-using-the-new-web-crawler-for-amazon-kendra/  
https://diptimanrc.medium.com/amazon-kendra-web-crawler-content-search-for-internet-and-intranet-401325c8b4ed  
https://community.aws/content/2eBsTWhvFFPUtzh2secQlNRBgta/prototype-a-rag-chatbot-with-amazon-bedrock-kendra-and-lex

Answer

You can consider to use web crawler as Riku_kobayasi said but  before choosing an approach, consider factors like the size and complexity of your website, the frequency of updates to the content, and your budget for hosting and computation costs. Additionally, make sure to comply with any legal or ethical considerations related to web scraping and data usage.

How can i make a AI chatbot with Amazon Bedrock that use my website data for answer the questions?

関連するコンテンツ