Frame 4735.png

This is a tutorial for creating a Dify LLM chatbot with RAG on Shakudo.

It will use Financial 10K as the knowledge base.

It will use Dify, one of the Stack Components on Shakudo.

Step 1: Ingest data into Dify

  1. Download the Financial 10K dataset from Google Drive

    Note that the data is also publicly available on https://www.sec.gov/edgar/search/

  2. Set up a model provider of your choice in Profile > Settings

    Note that you may need to provide an API key

  3. Upload the PDF files to Dify through UI and create a knowledge with the following configuration

  4. Give it a meaningful name, such as Financial 10K

    image.png

Step 2: Create a chatbot in Dify

  1. Create a chatbot with the following configuration

  2. Add the instructions below

    You are a financial data analyst who is more than happy to answer the user's questions.

  3. Add the knowledge you just created as the context

  4. Choose an LLM of your choice (e.g., gpt-4o)

  5. Test and publish your Dify chatbot.

    Don’t forget to publish your chatbot before leaving the webpage as they’re not automatically saved by Dify.

    image.png