Import documents in bulk
Import the pages that matter in your website
If a sitemap does not exist, have Libraria crawl your website
Import your Google Drive DocX + PDFs
Import your Youtube Videos from your channel URL
Import a Github Repo
Upload PDFs, DocX, and TXT files
At Libraria, a “Library” is synonymous with your “knowledge base”. It is where you store all the information and content that you want your AI chatbot to talk about.
For best results, it is important for you to keep it up to date and that the content will be easily digestible by our LLM / AI.
That’s why we give you multiple options to get the best results.
The most targeted and powerful way to answer a user’s questions. Use this if you want high accuracy results and your AI to answer a particular user’s question.
- When you elect to create a custom answer, Libraria does some backend work to improve the matching and ranking of similarity to the user’s question.
Scraping URLS, XMLs, and Crawling websites
Easiest way to upload thousands of urls
- for scraping any kind of URL, you have the option to turn on “AI Scraping”. AI Scraping will summarize and condense the most important information fetched from the webpage
- (Advanced) Alternatively, you can also specify with jquery selectors what we should take from your page.
- For scraping youtube videos, we take videos that have captions only. We then break it down to the timestamp so that when a user clicks on a source, they know exactly which part of the video it pulled it from
- Libraria does not save PDFs for you currently, but if you upload it with a separate URL, then when a user clicks on a source, it will take them directly to their PDF.
- If you want to be able to save PDFs, please contact us (here)[mailto:firstname.lastname@example.org]
- Everything else uploaded to your knowledge base is broken down into several “paragraph chunks”. These “paragraph chunks” are what is used to match a user’s input to the correct document you uploaded to your knoweldge base.