/LLM_Chatbot_DataTalks

Using Python and the Open AI API to create an AI chatbot that answers questions using a JSON file of course FAQs. A form of RAG.

Primary LanguageJupyter Notebook

Creating an AI Chatbot!

  • Part of the Data Talks Club LLM Zoomcamp
  • Connecting to the Open AI API.
  • The code and notes are contained in the Jupyter Notebook called 'ElasticRAG' saved in this repo.

The Task 📝

  • Using RAG (Retrieval Augmented Generation) to create a chatbot based on a document containing FAQs about various courses hosted by DataTalksClub.
  • The FAQ document is +25 pages long, so it is easy for users to overlook anwsers, or fail to look through the document altogether.
  • A chatbot helps solve this problem as it is an easy and intuitive way to find anwsers to questions quickly.
  • The chatbot works by searching for words in the question in the context given (the JSON file), it bases its anwsers only on this.

Tools & Libraries Used 🧰

  • Python
  • Open AI API
  • ElasticSearch
  • Docker
  • pipenv
  • JSON files
  • Jupyter Notebook

Example Output 📖

Here is an example of the Q&A between user and the chatbot! The user asks "I can't connect to postgres port 5432, my password doesn't work" and gets a sensible response from the chatbot 🏆

Screenshot of the example output of the chatbot

Resources and Shout-Outs 📚