Generative AI Foundations

1. Introduction to Generative AI
2. Large Language Models (LLMs)
3. Transformers architecture
4. LLM limitations (hallucinations, knowledge cutoff)
5. Use cases of Gen AI in industry

2. Prompt Engineering

1. Prompt design principles
2. Few-shot prompting
3. Chain-of-thought prompting
4. System prompts
5. Prompt templates
6. Prompt evaluation techniques

3. Introduction to RAG (Retrieval-Augmented Generation)

1. What is RAG
2. Why RAG is needed
3. RAG vs Fine-tuning
4. RAG architecture
5. RAG pipeline:
6. Retrieve → Augment → Generate

4. Data Preparation for RAG

1. Document ingestion
2. Data loaders
3. Text cleaning
4. Chunking strategies
5. Text splitting
6. Metadata handling

5. Embeddings

1. What are embeddings
2. Vector representation of text
3. Embedding models
4. Semantic similarity
5. Embedding generation pipelines

6. Vector Databases

1. What is a vector database
2. Similarity search
3. Indexing techniques
4. Popular vector DBs
5. Vector databases power semantic search in RAG systems.

7. Information Retrieval Techniques

1. Keyword search (TF-IDF, BM25)
2. Semantic search
3. Hybrid search

8. RAG Architecture Implementation

1. Building a RAG pipeline
2. Query processing
3. Retriever + generator integration
4. Context injection into prompts
5. Response generation

9. RAG Frameworks

1. LangChain

10. Building Real RAG Applications

Projects often include:

1. Document Q&A chatbot
2. Website knowledge assistant
3. PDF chatbot
4. SQL database chatbot
5. Customer support AI

11. Evaluation & Optimization

1. RAG evaluation metrics
2. Groundedness
3. Faithfulness
4. Latency optimization
5. Query caching
6. Prompt optimization

12. Deployment

1. Building APIs (FastAPI / Flask)

13. Capstone Project

Examples:

1. Enterprise document search assistant
2. Company knowledge bot
3. Research paper Q&A system
4. Customer support automation