Do you simple paste the entire content of a webpage inside a prompt? You prepend the question from an enduser with that content?:i.e. dear chatgpt api, could you answer the following question [user question] based on the following text [website contents]
I guess not. Probably an offline process where they scrape the websites into chunks and build embeddings. At query time first search for the relevant chunks and then put those chunks into the prompt? Would love more details though from the author!
Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
11–20 of 103 posts
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#12Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#13Sorry but I am not going to pay for this before trying it out on my content. Surely there can be a test function within the website which allows me to see what a user would experience?
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#14Congrats :) Question: do websites even have enough information to train on and create useful results?
This is an opportunity for the chatbot owner to add more relevant content to the website.
They don't have to think about what content to add. They can just see what questions the visitors have about their website and can add/edit their website content based on that.
After adding the improved content, retraining the chatbot with the new content will be as easy as clicking a button.
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#15Do you simple paste the entire content of a webpage inside a prompt? You prepend the question from an enduser with that content?:i.e. dear chatgpt api, could you answer the following question [user question] based on the following text [website contents]
Not exactly. Given one of my plans has 5000 pages, it is not possible to just paste the entire content in a prompt. Open AI API has a max tokens limit. I will first do some pre-processing on the content and fetch the relevant pieces of content before giving it as a prompt to the API.
Amazing concept btw - would love to see more examples (like a chatbot for a more well-known site).
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#16I want some intercom/crisp functionality there like getting an MS Teams chat started up with my customer optionally.
It's already in the plans.
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#17Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#18Sorry but I am not going to pay for this before trying it out on my content. Surely there can be a test function within the website which allows me to see what a user would experience?
Same here. I know it hurts to offer a free trial for something that already costs you money to serve (those API calls won't be free), but it's really hard to sign up without trying it first.
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#19This is really simple to build yourself. Just use SentenanceTransformers in python, pineconeDB, and whatever screen scrape library you line. Chunk your site content before storing in pineconeDB and then search and whatever you get back use as context for the GPT Chat API.
Some of the customers I got are non-technical people. So it was perfect for them.
They get a very good chatbot even if they don't know how to code.
Re: Show HN: SiteGPT – Create ChatGPT-like chatbots trained on your website content
#20Earlier quoted context omitted.
Not exactly. Given one of my plans has 5000 pages, it is not possible to just paste the entire content in a prompt. Open AI API has a max tokens limit. I will first do some pre-processing on the content and fetch the relevant pieces of content before giving it as a prompt to the API.
how do you decide what content on the page to index, and how to split them to fit the context window? Amazing concept btw - would love to see more examples (like a chatbot for a more well-known site).
When the user asks, I try to get the relevant bits and answer the question based on that.