Skip to content

Implemented an end-to-end Retrieval Augmented Generation (RAG) pipeline using Sentence-Window Retrieval and Auto-merging retrieval techniques to enable users to ask any questions about my work using an LLM.

Notifications You must be signed in to change notification settings

faizankhan29/RAG-For-Resume-and-Work-Portfolio-Analysis

Repository files navigation

Retrievel Augmented Generation (RAG) For Resume and Work Portfolio Analysis

This project is aimed at creating a Retrieval Augmented (RAG) pipeline to enable users to ask questions from my Resume and Work Portfolio. The user can ask any question about my career - work experience, projects, academics, accomplishments, skills and more.

In the attached notebook, we will follow three approaches to implement Retrieval Augmented Generation (RAG):

  1. Basic RAG Pipeline:
    • Basic Indexing of Documents using LlamaIndex.
    • Querying from the Created Index.

alt text

Source:https://learn.deeplearning.ai/building-evaluating-advanced-rag/lesson/2/advanced-rag-pipeline

  1. Advanced RAG Pipeline - Sentence Window Retrieval:
    • Creating Sentence Window Based Index (Breaks down documents into smaller chunks like sentences).
    • Querying most relevant chunks along with surrounding context.

alt text

Source: https://learn.deeplearning.ai/building-evaluating-advanced-rag/lesson/2/advanced-rag-pipeline

alt text

Source: https://learn.deeplearning.ai/building-evaluating-advanced-rag/lesson/2/advanced-rag-pipeline

  1. Advanced RAG Pipeline - Auto-merging retrieval:
    • Creating Automerging Retrieval Based Index.
    • Querying using auto-merging retrieval which merges information from multiple sources or segments of text to create a more comprehensive and contextually relevant response to a query.

alt text

Source: https://learn.deeplearning.ai/building-evaluating-advanced-rag/lesson/2/advanced-rag-pipeline

About

Implemented an end-to-end Retrieval Augmented Generation (RAG) pipeline using Sentence-Window Retrieval and Auto-merging retrieval techniques to enable users to ask any questions about my work using an LLM.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published